Hawker News

Typesafe AI raises $870M at $7.5B

typesafe.ai

372 pointsby tosh273 comments

armcat[125 comments hidden]
It's weird because two days after Jev was released there were a dozen decision models, a week later there are several dozen, mostly open source, OpenAI's own Decisions API [1] beats it, and you can easily finetune your own [2]. But as others have pointed out, this doesn't matter.

EDIT: As I wrote this Microsoft just released their own Decision-1 model [3].

[1] https://developers.openai.com/api/docs/guides/decisions

[2] https://unsloth.ai/docs/basics/train-your-own-decision-model...

[3] https://commandline.microsoft.com/microsoft-decision-1-model...

dkersten[24 comments hidden]
Most of them appear to be small LLM’s fine tuned for the role.

That’s a different set of properties in terms of size, cost, and latency. Jev (apparently, not like I’ve seen its insides) is extremely cheap, extremely fast, doesn’t cost any output tokens as it speaks the output natively, can’t get the output wrong because it speaks the format natively, and (presumably based on the docs), the context is separate from the question, meaning it should be immune (or at least highly resistant) to prompt injection attacks.

It’s not just about the accuracy of the result, it’s a collection of all the properties that make Jev interesting.

Jev took years to develop, I strongly doubt that a copycat that was put together within days after Jev’s release will be able to match it on a sun of its properties. Even if fine tuned LLMs can outperform it on raw accuracy.

devin[11 comments hidden]
Jev did not take years to develop. What it does was published in arxiv back in 2025. TypeSafe just marketed it.
baobabKoodaa[10 comments hidden]
What specific arxiv paper are you referencing here?
homarp[9 comments hidden]
https://arxiv.org/abs/2503.23303

and https://arxiv.org/abs/2510.01237

baobabKoodaa[7 comments hidden]
Yeah, I figured it was gonna be this one.

"SalesRLAgent: A Reinforcement Learning Approach for Real-Time Sales Conversion Prediction and Optimization"

Jev is a general-purpose thing. That is a specific-purpose thing. General-purpose thing is not the same as specific-purpose thing. What makes people think these are the same thing? I don't get it.

devin[6 comments hidden]
What do you mean? Jev is trivially different from what is described in this paper.
baobabKoodaa[5 comments hidden]
Bullshit. Below is copypaste from the paper in the section that outlines the "key contributions" of the paper. As you can see, it is focused on one specific problem: predicting sales conversions. So if you were to take this system and use it for some other task ("evaluate customer mood" for example), it would not work. Because, again, it is not describing a general purpose solution. It is describing a solution that is specific to one problem: sales conversions.

Copypasta:

• A reinforcement learning architecture specifically designed for sales conversation analysis and conversion prediction

• A synthetic data generation pipeline leveraging GPT-4O to create diverse and realistic sales conversations

• Novel state representation techniques using Azure OpenAI embeddings (3072 dimensions) with sales-specific features

• A meta-learning approach enabling the system to express confidence in its predictions based on conversation similarity to training data

• Integration mechanisms providing real-time guidance within existing sales platforms

• Extensive comparative evaluation demonstrating significant performance improvements over LLM-based approaches

rpdillon[4 comments hidden]
> Novel state representation techniques using Azure OpenAI embeddings (3072 dimensions) with sales-specific features

Jev is basically the embeddings side of an LLM. Yes, it's a good idea, but the moat is non-existent.

baobabKoodaa[3 comments hidden]
No. It can't simultaneously be both general purpose and having task-specific embeddings.
devin[2 comments hidden]
Hard to take you seriously, fam. The distance between these two things is trivial. The author of the paper implemented the "generalized" version of same in less than 24 hours. It's not novel, it is a continuation of what already existed in a boring way
baobabKoodaa[hidden]
The comment I was responding to describes Jev's significance as being specifically about embeddings:

> Jev is basically the embeddings side of an LLM

This is how the paper you are referencing is describing the "key contribution" as it relates to embeddings:

> Novel state representation techniques using Azure OpenAI embeddings (3072 dimensions) with sales-specific features

And now you are claiming that the author of the paper created:

> "generalized" version of same

It's a bit hard to guess what you are trying to say, but if I were to steelman your argument, I would guess that you mean: training a model with a vocabulary that has some tokens representing things like "OPTION_A", "OPTION_B" is in your mind the same thing as "generalized version of sales-specific features in embeddings"? Is this what you were trying to say?

zwaps[hidden]
This is a fine-tuned model. The author even states that the model is competitive with Jev only if fine-tuned on the evaluation at hand.

Literally misses the point of Jev, which you don't need to fine-tune to get accuracy nor - and no other model has this - some sort of out of sample calibration

hbrn[6 comments hidden]
> I strongly doubt that a copycat that was put together within days after Jev’s release will be able to match it on a sun of its properties

But why? If the simplest way to achieve Jev's capabilities (accuracy, cost, latency) is by fine tuning a small model, what makes you think that this isn't exactly what Typesafe did?

And even if they did something different - what makes you think it was a good idea in the first place, given how easy their results were replicated without any "secret sauce"?

dkersten[5 comments hidden]
My point is that it’s not replicated. You replicate the accuracy, but not the other properties. The Jev-competitors only proved that you can get or beat the accuracy, nothing about the other properties. Especially the “zero hallucination” output and the (if it works how the documentation make it sound) prompt injection resistant architecture. You can’t get that with a fine tuned LLM.
hbrn[hidden]
> zero hallucination

Plenty has been said about this claim. If you're still falling for this, I feel sorry for you.

If you remove wheels from your car, your car will get a "no speeding ticket" property, and yet there's nothing exciting about it.

> You can’t get that with a fine tuned LLM

Of course you can. All these claims are nothing but marketing.

user43928[2 comments hidden]
I understand most major model providers support passing a JSON schema that is strictly followed in the output, accomplishing the same 'zero hallucination' and prompt injection resistance.

The only difference I am aware of is that probabilities are better calibrated with these decision models compared to regular LLMs which can output hallucinated numbers where your schema allows a number.

dkersten[hidden]
I’m not claiming you can’t get some of the properties in some ways with LLMs, just that Jev is the sum of its parts, not just one or two properties.

LLMs can produce structured output by limiting the next token based on a grammar, that ensures correctly formed output.

But (presumably, again I haven’t seen Jev’s insides) Jev doesn’t need to do that, it just has to output probabilities for each answer, and can do it natively without artificially limiting output tokens.

So jev is cheap, fast, never produces malformed output, its output always carries correct semantic meaning (but this doesn’t mean it always answers correctly), and it separates correct from prompt (which if it truly does that is the most exciting part).

PunchyHamster[hidden]
Took me about 20 seconds to prompt inject the demo instance.... y'all would be great scammer targets
TeMPOraL[4 comments hidden]
Jev is something your favorite LLM could zero-shot months ago, if you pointed it to the right arXiv paper (some of which are linked in this thread).
mrinterweb[3 comments hidden]
That probably explains why there were so many competitors around withing days of the Jev announcement. They are not starting with a moat, and there doesn't seem to be any moat in sight. Just buzzword recognition because everything is comparing to "jev".
lifeisloving[hidden]
There will always only be a very small percentage of people who want to discover and build things, even with llms, its a very small subset of people though, and most people want off the shelf solutions.

Also Jevs purpose isnt to become its own thing. It will get aquired in 18 months by one of Andressen Horowitz's incestuous circle of companies and everyone will make money, and the person who buys it wont necessarily care if Jev itself makes them a ton of money. They're just passing chips around the table.

TeMPOraL[hidden]
Jev is the known idea, but with a slick website with grandiose claims and an impressive Doom demo on. It's also a catchy name - it caught on instantly, but mostly as a shorthand: it's easier to say "Jev" and "like Jev" than "classifier models" (or rather "<descriptive explanation of a specific shape of> classifier models").
creatonez[2 comments hidden]
Jev is almost certainly not immune to prompt injection
dkersten[hidden]
It separates context/state from prompt/questions. That’s a key ingredient for prompt injection hardening. If it is designed and trained to never treat state as instruction, and you only ever put untrusted user input into state, and never the questions, then why wouldn’t it be?

The jev documentation says this is what you should do. It doesn’t mean that it’s actually designed for prompt injection resistance, but it could be. We won’t know for sure until they either release more details or someone proves otherwise.

But the split is one that doesn’t exist in normal LLMs and it’s the exact split that is needed for immunity or resistance to prompt injection.

mlmonkey[2 comments hidden]
OpenAI's "Decisions" library has this in requirements:

To run the SDK examples below, use these OpenAI SDK versions or later: Python 3.26.0,

I thought Pythin 3.15.0 just came out, 3.26.0 must be really far off?

bayesianbot[hidden]
That is their Python SDK version, not Python version
scottyah[2 comments hidden]
> OpenAI's own Decisions API [1] beats it

Have you heard that from a different source than OpenAI? From what I'd heard other models haven't gotten close, and the open source ones are like running gemma4 E2B against Opus 5.5- sure, the API calls go in and are returned the same but the quality isn't close.

rockwotj[hidden]
I have ran a bunch of evals on production use cases where we are using Jev and OpenAI's model isn't close, but I'm hopeful about some of the OSS ones. IDK it's very cheap so I don't really feel like I need to look for an alternative. Why root for OpenAI over Typesafe? Isn't more diversity in the market good?
nlpnerd[hidden]
You are assuming that the VCs have done their due diligence. For a "hot" company like Typesafe AI, most likely little due diligence was done. That's the way it's played.
baobabKoodaa[4 comments hidden]
> you can easily finetune your own

no, you can't, and it's unclear why you would think this.

ricericerice[3 comments hidden]
you can easily finetune your own*

*if you have a sufficiently sized and quality dataset for the specific classifications you're targeting

baobabKoodaa[hidden]
And even if you do have that, you haven't made your own Jev, because Jev is a general-purpose thing, whereas what you have built is a specific-purpose thing.
9dev[hidden]
…and a team with experience and the hardware and the workflows to do this at scale, repeatedly to avoid drift, with proper infrastructure for evaluation and testing, and all of this work is somehow the core domain of your business.

Then yes, easy!

ttul[19 comments hidden]
But Jev established the branding and investors are betting that Jev will be acquired by one of the big labs soon - and if they aren't, the money itself can create a positive outcome by allowing Jev to hire incredible talent and scale the company rapidly.
Oras[18 comments hidden]
Rapidly? Their 2 years of stealth was replicated in 2 weeks.

I give it to them for creating the hype (good marketing), and for making a useful classifier. Not sure what they would scale rapidly though.

brink[hidden]
Either the investors know something we don't, or the market is irrational.
rsalus[2 comments hidden]
I feel like half the game now is marketing though, so I can see why they'd be attractive to an investor. Maybe if they scale they can come up with something.
SaltyBackendGuy[hidden]
Always has been imo.
ModernMech[hidden]
It’s the classic SV flip. You scale your investors, your executive team, your sales people, hire a bunch of engineering you don’t need, then sell the company. The company’s product doesn’t matter, the company is the product.
mococa[2 comments hidden]
> Their 2 years of stealth was replicated in 2 weeks.

Actually they stolen the idea from a paper.

shdh[hidden]
So did Oracle with relational databases by that logic
tiborsaas[5 comments hidden]
You can replicate any fitness app in no time and you will make close to $0. Brand recognition matters a lot.
throwaway7783[2 comments hidden]
100%. Product finesse + Marketing is now the "moat".
PunchyHamster[hidden]
"now"? It always was. Pre-AI we still had ton of great apps "losing" just because the couldn't be brought before the eyes of people that would like them
tyre[2 comments hidden]
Yeah but there are also network effects of using fitness apps. Every engineer I know switches easily between Codex and Claude Code when one gets better than the other.

Jev has been around for a couple weeks. Cost and performance matter more than anything. Staying with an existing provider (the # of people choosing Jev without already having a frontier API key is probably zero?) is way easier than this.

What the hell are we even talking about.

wordpad[hidden]
Enterprises are very sticky and they arent even chosing between claude and codex, many are looking at something like Kiro.
clickety_clack[4 comments hidden]
It’s like openclaw. There was a bunch of technically better ones that came along afterwards, but nobody remembers what any of them were called.
jeremyjh[hidden]
You say this, while Hermes Agent has been at top of the openrouter.ai leaderboard for several months and currently has 3X the token usage of OpenClaw.
real0mar[2 comments hidden]
What? OpenClaw has been completely replaced by Hermes and others in the discourse
Espressosaurus[hidden]
And Muse is what normal people use instead
OtherShrezzing[2 comments hidden]
Their 2 years of stealth involved a lot more than building their model.

People who can find things of value, and execute on those things, are worth investing in - even if that thing of value is replicated in short order.

Nobody was looking at decision models, but now that Jev has surfaced, everybody is.

vasco[hidden]
> People who can find things of value, and execute on those things, are worth investing in

Investing isn't binary. You can be investible but not at a 7.5b valuation.

doctorpangloss[hidden]
"It doesn't matter"

By all means, become an A16Z LP.

amelius[hidden]
I mean I'm already ditching my Apple stocks because soon AI will be able to replicate iOS and MacOS.
827a[33 comments hidden]
Decision models have the potential to have an even larger impact on the Real World than LLMs have to this point (which is obviously quite large). But the model itself matters less than the product experiences you build around the model, and its very likely that the incumbent labs are treating the area as something more like "oh yeah I guess we can ship that and then forget about it" rather than investing in what building business processes on decision models looks like. Unlike full language models, I don't think the primary business of Typesafe will be serving Jev at API pricing; it'll look a lot more like putting Jev at the center of a much more expensive suite of software.

There's the potential for an inverse LLM play. In contrast with LLMs, all that seems to matter is the model, and the products the labs build around the models are all really samey and boring; the same left panel list of agents, main view agent conversation, right hand extra context, and we're now in the era of everyone creating the same cutesey furry friend on top of all this tech.

nlpnerd[3 comments hidden]
Yeah, agreed. A model or primitive on its own has no moat and frankly limited value. The paradigm behind "System One" models on the other hand is potentially huge.

https://seldon-ai.com/blog/fronter-llms-are-semantic-interpr...

TeMPOraL[2 comments hidden]
Skimming the post, it seems to argue for reconstructing the very rigidity that LLMs let us escape, and that very aspect of LLMs is what made them useful and explode in popularity so much.
nlpnerd[hidden]
One may consider it "rigidity" but many application backends (including LLM powered ones) never needed the full unconstrained capabilities of an LLM. What people needed was the zero-shot capabilities over tasks as opposed to have to train a classifier, span model, etc for every task.

Constraints can be an advantage in many systems. Consider typed and untyped programming languages. Many advantages with typed languages over the latter in terms of development and efficiency.

porridgeraisin[3 comments hidden]
> Unlike full language models, I don't think the primary business of Typesafe will be serving Jev at API pricing; it'll look a lot more like putting Jev at the center of a much more expensive suite of software.

Precisely this. Should be top comment.

Also, the model moat is understated as training data for these purposes also accrues to the winner, which due to the first mover advantage as well as the distribution advantage you speak of, is typesafe. In contrast to relatively open coding data. Openai anthropic also have that, but like you say its a different business.

outofpaper[2 comments hidden]
The are still just 1tok output of pretty standard llms just along with the logprobs converted to some json
cheesecakegood[hidden]
At least in theory (TypeSafe has been pretty close-lipped about the details so this might just be hot air, and I think the evidence is a bit spotty) this is false, since they use a different reinforcement training method.

If the word “calibration” in probability doesn’t mean anything to you, the difference isn’t very apparent, but that doesn’t mean it doesn’t exist.

pantelisk[9 comments hidden]
Yes, the best way to think of a general classifier like this is like a smart switch statement. Essentially a "JEV" like thing becomes a sort of programming primitive. Once you see it, it's hard to not get excited.

But even if others surpass them and make better solutions, the fact that nobody was able to see it before typesafe is a testament of what they might be able to come up with next.

I sound like a fanboy but I swear I 'm unaffiliated with typesafe. I was building my own version of this way before they announced JEV (mine was ALE and it was mentioned here on HN for a bit), in use for VR gaming (so one can give commands to NPCs with voice and supports multiple commands in sequence in a single pass), but I missed the "killer usecase" of being a new primitive, like everyone else.

TLDR: There's a lot of value in thinking ahead and seeing the future. The clones are nice and exciting but they give me "I could have built this first, yes but you didn't" vibes. I hope they manage to keep it up and push the space forward again

aranelsurion[hidden]
I remember your blog post! Thanks for writing it, was pretty cool and a practical application.

Here if anyone is interested: https://pantel.is/projects/ai-gaming-companion/

overfeed[hidden]
> [...]the fact that nobody was able to see it before typesafe is a testament of what they might be able to come up with next.

Counterpoint: there are a lot of one-hit wonders, and they vastly outnumber the idea-factory people. This is not to minimize those people, a single idea can be very successful (see Zuckerberg), but it doesn't mean your subsequent ideas will also be great (see Zuckerberg)

skeeter2020[hidden]
>> TLDR: There's a lot of value in thinking ahead and seeing the future. The clones are nice and exciting but they give me "I could have built this first, yes but you didn't" vibes. I hope they manage to keep it up and push the space forward again

This all sounds intelligent and likely, and yet we can come up with countless counter examples where the first mover is not the big winner, and nobody cares about who did it first. There is typically way more value in nailing the execution of a big idea someone else came up with, rather than "seeing the future".

user43928[5 comments hidden]
Honestly, I don't feel the least bit of excitement here and I'm normally enthusiastic about AI.

Yes, this generates probabilities over a set of given options instead of the whole token vocabulary.

I don't see what's so exciting about it, compared to regular LLMs which support structured output options in the API.

senordevnyc[3 comments hidden]
You’re correct, it’s a great solution for a very narrow set of high volume classification needs that require very low latency. But that’s it.

I keep evaluating Jev for my product because of the hype, but the reality is that for my tasks, Luna is more accurate, only a little more expensive, and the latency doesn’t matter. I’d rather spend the extra $50 / month or whatever than have to shoehorn in another API and provider, and also lose the ability to change reasoning level and get reasoning summaries for eval purposes.

jeena[2 comments hidden]
Did you look into self-hosted open source decision models because they are quite capable also much faster and much easier to integrate and you don't need to pay anything.
senordevnyc[hidden]
I keep hearing Jev is better for accuracy, and still in my use cases Luna beats it. Latency doesn’t matter to me. The cost for additional Luna tokens is negligible for my use case. Idk, I just can’t figure out where to use it instead of an LLM.

But that’s just for my product.

pantelisk[hidden]
LLMs have higher accuracy and are more versatile, but that comes with a cost (slower and more expensive to run). On the other end you can have your own mini classifier something built on top of ModernBert or DIET or even a hash classifier (these are narrow).

Models like JEV fall in the golden middle. Somewhat cheap, somewhat fast, somewhat general.

Many people have replied saying that narrow classifiers (eg bert) are better, and they are better in many ways (essentially free and faster). But... the real world is messy, in my experience it's actually very hard to make a really good classifier that will work across a specific niche domain especially when there is very fuzzy input. And the most interesting real world scenarios are nuanced and overtime there will be edge cases discovered where it fails and that will require retraining and re-evaluation etc etc. It can becomes a full project on its own that eventually collapses into a game of whackamole (improved in some direction but regressed elsewhere). Things like JEV (or similar models) provide ease of mind, just off load the complexity to it and move on to the next challenge type of thing.

nico[4 comments hidden]
> But the model itself matters less than the product experiences you build around the model

This is a very important insight. And it applies to LLMs as well. Very few people were impressed with the capabilities of GPT 3, it was mostly a techie novelty

But then when they added chat on top of gpt 3.5, all of a sudden it was a huge hit. Sure there were improvements in the model from 3 to 3.5, but the biggest impact was from the chat experience

Conversely, when they created Eliza, a basic chatbot more than 50 years ago, people even got addicted to it, despite it’s ai model being something super rudimentary and basic compared to what we have now. The model capabilities didn’t matter as much as the experience the chat created

vrighter[hidden]
i read that quote as "how it looks is more important than if it works"
Darmani[2 comments hidden]
The instruction tuning between 3 and 3.5 was a big deal. If you prompted raw GPT 3 "Tell me a story about a mouse", it was as likely to continue "2) Tell me a story about a cow. 3) Tell me a story about a rabbit" as it was to tell you a story. I tried to use GPT 3 for work and found its only utility was to get over blank page fear by writing something so bad I would fix it in anger.
dannyw[hidden]
It was impressive, but not that much. If you had a basic prompt prefix like:

“The following is a conversation between a human and a helpful AI assistant.

Human: [prompt]\n AI:”

You would get consistent conversational chat, with the main limit being the model’s limited context window, and obviously frontier intelligence at the time.

I played around quite a bit with davinci-003 and conversational systems before ChatGPT. I kept using this format and API for a while, but at least during the ‘free research preview’ era, found it generally smarter than chatgpt. 3.5-turbo was probably the watershed moment where I moved away from completions on base models; to chat completions.

If you’d like to emulate this experience, spin up a base (not instruction tuned) checkpoint of Llama 1; or a more recent base model. You might underestimate how much ‘intelligence’ you can get :)

Back then the API exposed a lot of controls from sampling to logits, so you could specify a custom stop token (some rarely used Unicode); then switch to near-greedy sampling for more reliable “function calls”, etc; and then switch back to sampling for chat.

In the early days, before releasing something with the API, you had to get your use case approved by OpenAI, like the App Store review process.

olalonde[12 comments hidden]
> Decision models have the potential to have an even larger impact on the Real World than LLMs have to this point

Why?

porridgeraisin[5 comments hidden]
If youre the one or have spoken to someone implementing "AI solutions" inside large companies recently, a decent chunk of it is soft policy enforcement with very basic context. They moved from gemini 2.5 flash lite type models to jev. Which is also why I found the price comparisons to "GPT Astra" on Twitter rather funny.
zhivota[4 comments hidden]
Right, example being expense pre-approvals (for low dollar value purchases like office equipment). General purpose LLM not required, and the product experience surrounding it does matter (reporting, auditing, fine tuning future responses, etc.).
majormajor[3 comments hidden]
A lot of that seems like low-value-add/low-margin "simple" automation.

I haven't personally yet found a net-new-capability bigger than that from LLMs here.

sks_15[hidden]
It really shines when you use that decision routing in industries where they need to automate these routings reliably with very low latency, higer certainly and ,more important, at scale. Take call centers for example. If a voice company needs to automate escalation etc. an llm would be an overkill. That's where system 1 models shine.
porridgeraisin[hidden]
It is. But the thing is, that is also a lot of the usage for "AI" these days. LLM based search pipelines are being deployed but not really used. Chatbots are a great hit. And then of course, in coding, finance, sales, accounts and audit, etc, it is being used to accelerate output.

These guys also love text to sql. I know 4 people each implementing text to sql for their separate companies. Or text-to-redash dashboard in some cases. Funnily enough, text to sql is one of the things where LLMs are clearly sub-human. Mostly due to lack of easy verifiability.

reexpressionist[2 comments hidden]
A major limitation of directly using "System 1 decision models" (a.k.a., logistic regression, and related uncalibrated classifiers over the output/logit space) in enterprise settings (or other high-stakes settings) is that such estimators are not reliable estimators of the predictive uncertainty in the presence of covariate shifts, and such estimators also lack a means of instance-wise data attribution (i.e., interpretability-by-exemplar), so they're not the ideal estimator for verification, routing, uncertainty over retrieval and tool-calls, etc.

For decision-making with neural networks, we instead need the older idea of estimators of the predictive uncertainty with constraints in the feature-representation space (over training/support of the estimator), as with Similarity-Distance-Magnitude estimators: https://pypi.org/project/reexpress-sdm/

SwellJoe[hidden]
That's a lot of words to not answer the question.
rebyn[4 comments hidden]
Is it me or so far all the current HN replies to this particular “why” ask seem not even remotely answering it?
SalariedSlave[hidden]
it's not just you.

the claim is extraordinary, yet the answers cover only the mundane.

iinnPP[hidden]
I just wanted to point out that I have started to note this "phenomenon", seemingly globally, on any topic.

The sum of people not even close to the topic has risen, according to my feels at least.

Has this rang true for anyone else?

PunchyHamster[hidden]
Because answering the why reduces investor hype train
baxtr[hidden]
I think you aptly describe the benefits this approach can offer.

What I haven’t seen answered is why it isn’t reproducible within very short amount of time and low effort.

Sure, the large labs are the lazy incumbents at this stage. But any other startups would be able to copy the experience. What’s the barrier of entry that I am missing?

moralestapia[hidden]
Nothing beats nepo, brother.
hkalbasi[hidden]
> OpenAI's own Decisions API [1] beats it

Jev is 42$/B but OpenAI is 100$/B token.

bushbaba[hidden]
A major VC could type safe ai money, then head to a larger AI company looking to raise their series E+ and demand they acquire typesafe as part of their funding allotment.

such an arrangement can end up beneficial to the VC firm

throwaw12[hidden]
You are right in terms of how fast competition created alternatives.

But, for OpenAI this is not a primary business, for open source models as well, so they will not be chasing the market and customers to buy their product and promise them to maintain it.

TypeSafe will do all this, they will try to understand your use cases and then solve your pain point, while others are providing raw material.

jgilias[10 comments hidden]
Isn’t the OpenAI decisions API basically just Luna cosplaying a decisions model and pretending the confidence score isn’t just a hallucination?
phalangion[hidden]
What’s the difference?
hbrn[8 comments hidden]
And what do you think Jev confidence score is?

Here's a hint: confidence is not generated by a model.

jgilias[5 comments hidden]
Thanks, fixed my understanding!

Do you think though that Luna being a model post-trained for chat produces over-confidence in logprobs?

hbrn[4 comments hidden]
Yeah, but I wouldn't be surprised OpenAI's decision API is a post-trained Luna with confidence calibration.

Typesafe claims that Jev is calibrated, but there are plenty of examples where it completely fails (predicting die roll being the most obvious one).

Unfortunately calibration is hard to benchmark.

shados[3 comments hidden]
The dice roll prediction is about the way the prompt is setup misunderstanding how Jev works (they treat the confidence score as a probability score, which it isn't).

If you instead give it a list of probability for each number and ask it whats the probability of each number, the result will be accurate.

hbrn[2 comments hidden]
> give it a list of probability for each number and ask it whats the probability of each number

Did i hear that correctly? In order for Jev to be accurate you have to give it the answer before asking for the answer?

(btw this is exactly how Jev is playing games).

jgilias[hidden]
That’s circular, sure. But if we think about potential real-world tasks where someone might, say, use it to classify on “does this advice correspond to our policy docs”, you’d absolutely push the answer (the policy docs) into its context..

What am I missing?

As in, neither LLMs nor Jev are truth engines. Truth comes from the provided context.. Plus weights.. kind of fuzzy, sorry I’m thinking out “loud”

adrian17[2 comments hidden]
Maybe I'm missing something, but why couldn't it be generated by the model? In older classification tasks with transformers like BERT, you could absolutely obtain a confidence score.
hbrn[hidden]
Jev API returns both confidence and probabilities.

But confidence value is just a function applied to probabilities. It is not coming from the model, and it carries no additional information.

It is documented btw, and yet you will see plenty of claims that Jev is better than LLM because it returns both.

girvo[hidden]
Counterpoint: my work has already allowed us to call and test Jev. Those others? Who knows when, if ever.
user3939382[hidden]
Investments aren’t made because the product is amazing, they’re made because there’s a compelling exit scenario. Engineers don’t want to hear this but more generally, the critical success factors for a business aren’t product or engineering they’re relationships i.e. sales and team dynamics. If technical excellence dictated business outcomes in tech Salesforce wouldn’t exist for example.
nico[hidden]
Yup, I also released an open source classifiers tool, Jeffy. It comes with 68 pre trained classifiers which run and train on CPU alone. They run locally and are faster than Jev/Laya/Decisions. And they can do things like label email, all the way to even playing Doom

* https://jeffyclassify.com/

* https://playground.jeffyclassify.com/#doom

* https://github.com/nicobrenner/jeffy

rpdillon[hidden]
In my experience with OpenAI's decisions endpoint, it tends to return either 0 or 1 and doesn't return middle confidence levels very much at all. Would be interested to hear if others have experienced the same.
shados[2 comments hidden]
I honestly was worried for them. Now, even with all the clones, they generally still come up on top in price and latency, but "good enough" is often sufficient.

Guess the (investment) market has spoken.

nacs[hidden]
Hm? Jev's actual (network) latency is not that great and even small local models are doing far better latency-wise.

The Microsoft article GP linked even shows the MS model having 95ms latency.

Even on price Jev being matched (the same MS model is "Input tokens cost $0.042 USD per million tokens. Output tokens are free.", same as Jev).

Cloudflare's Clef-flash model is actually slightly cheaper: "$0.038 in / $0 out per 1M" too.

Jev is being matched or exceeded in performance and price within a month of them going public.

shmoogy[5 comments hidden]
Jev outperforms clef and OpenAI decisions on most of my tasks, outside of when I need multi modal image going into it. Jev is also cheaper but costs are minimal overall.
lifeisloving[3 comments hidden]
I just dont understand why anyone needs a general purpose classifier. Just build a classifier for your specific use cases.

Fortunately Jev is cheap, so I dont think it matters too much, but I think its robbing people of the opportunity to learn and implement this themselves.

Also, I dont really want 3 companies responsible for censorship/classification.

jofzar[hidden]
That's why it's useful, it's for when you don't know what needs to be classified or if you want to offer this classification models to your users.
nlpnerd[hidden]
Because training your own model requires substantial effort around data annotations, and then dealing with data drift etc in production.
nwienert[hidden]
Hm I had built out a system that tested a few hundred decisions and clef beat Jev pretty handily, I think there was actually 0 instances where clef regressed and it improved about 20% of the false decisions.
charm137[3 comments hidden]
I see Typeface AI have 30-40 people working for them on LinkedIn (some are VC advisors / board members), a lot of them being engineers. Their openings suggest a strong developer market focus (to begin with). I don't necessarily see a lot of people at the company with enterprise sales channel experience, but they already claim to have several Fortune 500 companies as clients (maybe the VCs are helping there or there is natural dev-driven traction).

So they clearly have a product, a strategy around it and perhaps the compliance scaffolding (SOC2 Type II etc) that may be needed before actually being able to charge money for it. They also have the right brand names associated with the founding team. Execution, so far, seems good enough to create a splash, at least.

As an investor, the question(s) to ask is (in my view): "How do they make money? Will that way to make money survive?". The answer to the first: selling input tokens and perhaps subscriptions/credits eventually.The answer to the second: "Yes, but with the risk of unit revenues declining faster than their unit costs". How can they mitigate this problem: by being big (scale / mindshare etc) so that their unit costs (including for customer acq) fall faster than their unit revenues will - I believe that is the question most AI companies are trying to tackle these days. Any new competitor will have to tackle basic fixed costs (of getting started) first before even getting to the stage of having the luxury of worrying about unit-economics.

So yes, they might eventually be competed away but whoever is in their team is trying hard to make a useful product/ecosystem and that should be applauded, not ridiculed with "it's all marketing". This is way more than a simple github/huggingface-based open-source replica solution can hope to achieve without institutional backing (either big-tech or system-integrators).

What should rightly be questioned, of course, are the valuations the VCs are providing to them in hopes of passing this hot potato to a willing buyer (say a hardware maker like NVidia) - the incentives there are very well defined and depend very much on perceived TAM (which lately is on very shaky ground given how far token pricing has fallen causing, among other things, OpenAI to "miss" on the market's expectations for annualized revenues, even before they're listed!) [1]

[1]: https://www.ft.com/content/b66a9858-f8fb-46cb-b506-44bfe26fc...

notfromhere[hidden]
An Enterprise client at that size can just be someone at an f500 put a credit card in
9dev[hidden]
Currently you can’t even add a VAT ID to your account, leading to invalid invoices for all EU customers.

So yeah, not too much enterprise sales experience there for sure.

neilellis[2 comments hidden]
Did you actually try them, I tried a sample they were rubbish compared to Jev - I'm not saying they will survive, but the competition CURRENTLY sucks - and also they aren't actually cheaper if you need them at scale. OpenAI is coming in twice the cost and so is Cloudflare. Ad yes you can finetune, and yes I do, but it also SUCKS. Fine-tuning and managing your own datasets is yet another timesink. So we'll see, but it's not quite black and white.
phoghed[hidden]
Yeah, tried a couple open source ones, same experience. Seems to me a ton of people are in a mad dash to create one of these and most seem to have over fit on the Jev benchmark to get some claim to fame of beating them or replicating their couple years of work in two days or whatever.

Then everyone breathlessly repeats this story about how Jev is useless because open source models they never tried claim to do the same thing and better.

fastball[hidden]
And there are dozens of open-weight LLMs, but OpenAI and Anthropic are both immensely valuable because their models are actually intelligent.
jofzar[hidden]
Testing I have seen from internal teams have shown it's still the best and cheapest model so far for pure classification work.
vrighter[hidden]
because all that is needed is to delete the outer while loop. it's an llm that is only run once and generates just one token. But it outputs confidence levels! Yeah confidence levels are the only output from an llm
Ozzie_osman[2 comments hidden]
It's possible that Jev has more in the pipeline. I mean, for shipping something so paradigm-shifting, this would be a bet on the founder/team continuing to do things faster or better.
c7b[hidden]
Paradigm-shifting? A classifier? It's great and it's a cool use of Transformers, but AI really didn't only start with LLMs. We used to have an own term for AI models that can handle text instead of just numbers, now a model that outputs numbers instead of text is considered paradigm-shifting.
fennecbutt[2 comments hidden]
Because the American stock market is all overvalued stuff with monopoly money.
vrganj[hidden]
Marx called this fictitious capital.

> Fictitious capital could be defined as a capitalisation on property ownership. Such ownership is real and legally enforced, as are the profits made from it, but the capital involved is fictitious; it is "money that is thrown into circulation as capital without any material basis in commodities or productive activivity".

https://en.wikipedia.org/wiki/Fictitious_capital

christina97[24 comments hidden]
Everyone appears surprised by this news. It’s clear that they don’t have a product with some incredible moat. But they clearly have good engineering and product people that came up with a product people wanted. On top of that they have very strong marketing muscle that took the AI world by storm. And as far as I’ve seen, they still lead in some part of the latency-quality (-cost) curve?

They may well be a good team to throw money behind if you are hoping to bet on a new AI lab.

soleveloper[8 comments hidden]
They don't lead on latency nor quality; but their execution was superb
tietjens[6 comments hidden]
Who they trailing on quality?
baobabKoodaa[4 comments hidden]
I'm not the person you're asking, but:

https://benchmarkheaven.com/jev-models

According to this benchmark, Jev is currently trailing Quyet-1.0-Large and a few other hastily put-together LLM-based decision API-like setups.

tpetry[3 comments hidden]
And the 'better' ones are slower and cost more for a tiny bit more accuracy. Its hard to sell that as being better when speed and price have been JEVs main selling points.
baobabKoodaa[hidden]
The top alternative right now lists speed as faster than Jev?
senordevnyc[hidden]
The space of problems that require super low latency AND are inputting so many tens of billions of tokens that the price matters, is a tiny one.
deepsquirrelnet[hidden]
Here's another leaderboard: https://huggingface.co/spaces/multimodalart/jev-decision-ind...
nwhnwh[hidden]
What even is this? What does it do?
verdverm[hidden]
Wonder if they can get coin flips and dice rolls to make sense with this fresh funding, or if it even matters to people.

I have no faith in the technique if it cannot do the basics (i.e. not real probabilities, the confidence for coin flip outcomes)

tried it a couple of days ago here: https://jevplayground.com

the "not real probability" disclaimer only appears after you get a result

rvz[3 comments hidden]
This is all due to 40% marketing, 50% execution and 10% credentials (with the founders being associated with creating ChatGPT).

If anyone else came up with the same concept on a Reddit thread (they have) it no-one would care without those characteristics even if you are "first".

Rebranding, execution, marketing, ex-<big_name_company> and mostly importantly, hype is what gets the investors scrambling into throwing money at you.

objektif[2 comments hidden]
More like 10% 10% 80%.
JamesSwift[hidden]
35% 10% 55%
binlog[2 comments hidden]
The model itself is a negligible part of the valuation. The company is priced as an acquisition target.
redanddead[hidden]
Every startup is priced as an acq target
hbrn[5 comments hidden]
> people that came up with a product people wanted

We've yet to see whether this is true, or is it just manufactured demand. There are dozens of Jev demos, but pretty much all of them are either cool but useless, or simply fake (i.e. harness doing 99% of the work).

vanuatu[4 comments hidden]
since their release they surpassed 100M arr and 1/3 of the F500

its clear they own the mindshare around this type of primitive which is a massive premium

hbrn[3 comments hidden]
Measuring ARR based on 7 days of data is beyond stupid.

You can create a company with 2B shares and sell one share to your friend for $1000. Lo and behold, you own $2 trillion dollar company, leaving Elon behind.

qlte[2 comments hidden]
Especially for a company still in their 15 minutes of fame that could very plausibly turn out to be a one hit wonder and a trivia question in two years.

Multiplying out revenue from the peak of their mini hype cycle while a dozen well financed competitors target them directly seems extremely optimistic.

hbrn[hidden]
Ha, two years is too generous.

If they weren't lying about $100M ARR, they had to be at 6.5T/day for a week to hit that number.

They also claimed they were at 1T/day just couple weeks ago. That's $15M ARR.

85% drop in just a couple of weeks?

cootsnuck[3 comments hidden]
> On top of that they have very strong marketing muscle that took the AI world by storm.

Not them per se, but doomers.ai [0] which is a uh... "launch virality agency".

So less that Typesafe has a strong marketing muscle, and more that they paid at least $100K [1] for "organic" buzz.

[0] - https://doomers.ai/work/typesafe-ai-case-study

[1] - https://doomers.ai/guides/ai-marketing-agency

asa123[2 comments hidden]
you’ve got to be kidding me. i know it ISNT fraud but i can’t help but feel like this SMELLS like fraud
adrianmsmith[hidden]
Why?

I read the attached links and it just looks like a marketing agency?

solatic[hidden]
A great team is now worth $7.5B?

They can be both great and over-valued at the same time.

prometheus1992[50 comments hidden]
I really don't understand how this can be. I have sat in fund raising meetings with VCs in toronto and my experience is that there is shit ton of due diligence at the tech level. a product which has no moat, was already available, was duplicated within a couple of days is valued at 7B - i thought we were past the peak of the hype cycle.
slopinthebag[hidden]
brb gonna wrap claude with a new form of prompting and raise 500 mil
Onavo[hidden]
You are used to dealing with companies where the money bags hold the power.

When you are in the middle of a boom cycle, it's the hottest company that has the advantage. Investing in them is a matter of privilege and they get to pick and choose.

Also, Canadian VCs are bottom of the barrel as far as VCs go.

besterman23[3 comments hidden]
Probably the “nobody ever got fired for buying IBM” effect. If there’s a use case for the tech, buying the most well known implementation of it will always be useful for people who want credit without the threat of blame. This funding is based entirely on the hype and a bet that TypeSafe will have name recognition.
johnfn[2 comments hidden]
Isn't this the exact opposite? When people were saying that saying, the connotation was that IBM was an old, stodgy company that had been around forever. (These days I often think "No one got fired for choosing AWS"). Typesafe is a hot new startup that could, to my eyes, easily burst into flame or die in the next year.
besterman23[hidden]
I’m saying the bet is in them becoming THE System One Model company.
kingcauchy[hidden]
It might be the people that are being acquired too, at huge inflated ai researchers salaries.

Acquired in the vc sense… not literal exit.

reticulates[2 comments hidden]
The lack of a “moat” is mostly irrelevant because success is not decided by who can or can’t be cloned. TypeSafe invented[1] a new approach that became wildly popular almost immediately, if they can do that once, they can probably do it again. Venture capital is big bets, of course TypeSafe is going to fail, that’s inevitable, but if it has even a 10% chance of capturing 1/10th the market cap of OpenAI then it is a great investment! Plus, money means nothing any more, they’ve raised less at a lower valuation than Instinct, a personal assistant.

[1] not really but they did some innovative things and popularized a concept

nateb2022[hidden]
I'd guess they justified the funding by revealing some grand scheme for a new product that they just need more runway to produce.
phren0logy[2 comments hidden]
The API was duplicated, the results were not.
JamesSwift[hidden]
I mean, a large majority of "pro jev" posts I see purely focus on the speed/cost and take their word on the accuracy. The entire product is "we guarantee the response will fit this schema, and will do it fast and cheap". Personally, I dont see how anything in this class can be "production ready" as a general classifier without the ability to fine tune on top.
dvrp[hidden]
That is not how it works in the US.
InsideOutSanta[hidden]
> there is shit ton of due diligence at the tech level

Maybe in some cases. But counterexample, courtesy of The Information:

"It took just 15 minutes for Blue Owl executives to agree to invest up to $10 billion in future projects alongside real estate firm Primary Digital Infrastructure during their first in-person meeting two years ago, said Primary chief investment officer Bill Stein."

https://www.theinformation.com/articles/blue-owl-eyes-new-de...

AI seems to make some people lose their damned minds.

nlpnerd[hidden]
This is the difference between a "hot" company in a good ecosystem like SF. Yeah, the funds can take 2-3 months to do their due diligence. By the time it's done, the round has closed, and then what good is the due diligence?
bix6[hidden]
They got money and it needs to be put to work!
baobabKoodaa[19 comments hidden]
> was already available

no, it was not

> was duplicated within a couple of days

was it already available or did it become available in a couple of days? it cant be both (neither is true, actually)

hirako2000[18 comments hidden]
The tech behind it existed. They made it a specific product, and got replicated in days.
baobabKoodaa[17 comments hidden]
Unclear what you're referring to. Please stop making vague claims and be specific.
euleriancon[hidden]
I think it is clear he is referring to zero shot classifiers with an LLM backbone. That tech has existed for a long time.
prometheus1992[15 comments hidden]
maybe you entered the AI space during the vibecoding era but there had been ton of useful models before that. especially zero shot models- both for text and images.
baobabKoodaa[14 comments hidden]
Everything you said here is false. No, I didn't enter the AI space during the vibe coding era. I was training custom ML models back in 2017. And no, there haven't been models comparable to Jev before Jev was published.

Jev is:

- accurate

- general purpose

- fast and cheap

Models we had before Jev had at most 2/3 of above qualities, but none of them were 3/3.

hbrn[7 comments hidden]
> Models we had before Jev had at most 2/3 of above qualities, but none of them were 3/3.

You're the one being deceptive here. Jev is trading accuracy, speed, and cost for generality. It's less accurate, slower and more expensive than trained classifiers. So it's still 2 out of 3, but with decimals. Maybe 2.2 out of 3 if I'm being charitable.

And the reason we didn't have that before is because nobody thought it's a good tradeoff.

baobabKoodaa[6 comments hidden]
When you say "trained classifiers", you are referring to models which are trained (or fine tuned) to work on one specific problem, right? That is the opposite of "general purpose".

Would Jev be more accurate in a specific task if it had been developed only for that task, as opposed to general purpose? Of course it would. So, sure, Jev is trading accuracy for generality. According to you "nobody thought it's a good tradeoff", which again is false, there was huge demand for a cheap and accurate general purpose classifier.

hbrn[5 comments hidden]
> work on one specific problem, right? That is the opposite of "general purpose".

A business doesn't need Jev for the sake of Jev. Most business are solving specific problems.

And fine-tuning got a lot cheaper these days - I've seen claims here on HN that ~500 examples is enough to beat Jev.

> "nobody thought it's a good tradeoff", which again is false, there was huge demand for a cheap and accurate general purpose classifier

There wasn't. The hope is that there was a latent demand, but we've yet to see if it's truly latent or just manufactured.

Noone is saying "hell yeah, finally we got a general purpose classifier, my business needed it so much". The typical message is "this seems cool, let me see where I can apply it".

The fact that name itself is a play on Jevons Paradox illustrates that there was no demand until Jev was released.

baobabKoodaa[4 comments hidden]
Yes, businesses are solving specific problems, but most businesses have more than 1 problem to solve. No, it is not economical to pay a data scientist to develop a custom model for each of your tiny problems. It is often much more economical to use a general purpose solution, like an LLM, or now, Jev.
hbrn[2 comments hidden]
At this point there's no need to pay a data scientist. You can literally ask Claude to do everything for you: extract real examples, classify them, post-train a model, and ship an API.

Now, it could be viable if your business has literally hundreds of problems thats require classification. I just haven't seen those.

I treat the fact that almost noone was doing that as evidence that decision models aren't that useful/groundbreaking. That, and the fact that every single demo I saw was either fake (e.g. playing games), contrived, or plain wrong (e.g. using Jev for compaction).

baobabKoodaa[hidden]
No, we're not at the point where you could ask Claude to do all of that, unless you have a super easy problem to begin with and/or you don't care about output quality. Feel free to link a counter example.
majormajor[hidden]
Are there enough "tiny problems" for businesses in the world to justify this sort of money?

The appeal and claim of LLMs for businesses is solving big problems.

LLM valuations to solve tiny problems seems iffy.

--

Are there significant problem domains LLMs are bad at that Jev is good at? Vs just 'Jev can do a subset of LLM things faster/cheaper'?

prometheus1992[2 comments hidden]
>>I was training custom ML models back in 2017

but you truly do sound like an angry 19 year old from your arguments.

- accurate - on what? on trust me bro benchmarks?

- zero-shot model are fundamentally general purpose.

- fast and cheap ; models on hf are FREE and fast enough.

baobabKoodaa[hidden]
"Models on hf" (unspecified) are "FREE"? Like "free to download"? Sure, but nobody was talking about that. They cost money to run inference on. Unless you are talking about some tiny toy models that are useless for any non-toy problems. You clearly don't have any idea what you're talking about. Just stop, man.
gpugreg[4 comments hidden]

    > I was training custom ML models back in 2017.
Maybe you are a good person to ask my question then. I have not looked into Jev much, but is it much different from using a regular LLM and constraining its token output to the action space? (e.g. like using llama.cpp's GBNF grammars). Is it just that Jev's "confidence scores" are significantly better than the softmaxed logits? Or is there something else I am missing?
baobabKoodaa[3 comments hidden]
What you're missing is: cost and speed. Otherwise, it's very much like running an LLM and constraining the output.

I don't know how good Jev's "confidence scores" are, but I would be surprised if they were in any sense better than logits from some good LLM. One advantage of Jev here is that the confidence scores are easy to access. Most LLM API providers don't provide an easy/convenient way to access the logits. But that's a minor point, you could of course build something like this with LLMs (and many people have).

gpugreg[2 comments hidden]
llama.cpp is already extremely fast for single-token responses (<5 ms). I can't see Jev being faster when taking network latency into account, except maybe for multimodal inputs.
baobabKoodaa[hidden]
Sounds like you are running a tiny toy model if you can get generations in under 5 ms? Typical response times from LLMs for typical "jev-like" queries from OpenAI and Anthropic are 2s-10s. Not milliseconds. Seconds. Same queries from Jev are like 0.2s. and the cost is 1000x.
geoffschmidt[hidden]
There is a belief that there is going to be at least one more breakout success in startup AI labs - rather than OpenAI and Anthropic being the final word - and so investors want to own a part of whichever companies seem most likely to be that success. If you start from that premise and stack rank what company that might be, you could quite reasonably put TypeSafe toward the top of that list right now, based on the people at the company and the ability they've demonstrated to ship stuff that people care about and cut through the noise in a crowded space.

Also the situation isn't static. Investors know that the act of writing them a $870M check itself increases the chance that they'll be one of the winners, because that will attract more talent, customers, and funding to the company in a self-reinforcing cycle. And investors know that other investors know that, and that someone is going to write them that $870M check, so to some extent they're forced to think of the company as having already been successful at the fundraising and already having that momentum boost.

Only a small number of investors in the world can play the game at this level, because you have to smart enough to be right (often enough), and you have to be established enough to see the deals (be on every CEO's short list - because CEOs are only going to seriously pitch 5-10 VCs on a hot deal, if that). Otherwise you can't pull it off. Martin Casado and his team are among the few that can and I think their results reflect that.

dist-epoch[6 comments hidden]
What was the moat of Dropbox? Of Instagram? Of GitHub? Or of countless other very successful startups when they started?

Anyone know when this "have no moat" meme appeared? Even 5 years ago I don't remember seeing it on every post.

hirako2000[hidden]
Whether a product has network effects makes a difference.

I would argue Dropbox did have a moat. It didn't merely store your data. It made it possible to make backup efficiently when bandwidth wasn't all that good.

Reading "no moat" so often is also tied to the fact those companies happen to be getting surreal valuations, at a quite early stage, showing no profit, building a tech that doesn't seem difficult to reproduce.

IshKebab[hidden]
Dropbox had a super smooth UX that somehow nobody else replicated (seriously Google wtf). Instagram and Github won on network effects.
uuue[hidden]
Instagram clearly did have one - Zuck had to acquire after internal efforts failed.
JamesSwift[2 comments hidden]
Dropbox: is I have to move my entire collection and setup the app on all my devices

Instagram: I have to move all my posts and also convince all my network to move over

Github: not a terrible example

If I decided to move off of jev tomorrow it would be an api key and a base api path update. Maybe 30 seconds of work.

dist-epoch[hidden]
I meant more what was their moat when they launched, compared to their clones.

You can say the same thing about OpenAI/Anthropic, just an endpoint update, maybe a harness change if you use the CLI.

People have in fact been saying "what is the moat of OpenAI/Anthropic", yet here we are, trillion dollars valuation.

cmrdporcupine[hidden]
Yeah your problem and my problem and others around here is the word you just said there... "Toronto." Canadian investors are risk averse as hell. And cheap. They can make more money helping sell bitumen or real estate, why bother with arcane tech?

And if you could put the words "Bay Area" or "Stanford" or "San Francisco" next to your name... different story.

The VCs are not buying the idea or the tech, they're investing in the people. And they invest in a formula that has already worked for them before to make big coin. Prop somebody up, let them hire like crazy, and then get them get acquired, and then cash out. They don't care if it fails if they can make it succeed 1/200 times.

Canadian investors want you to have already succeeded before they help you succeed a tiny bit more.

fidotron[3 comments hidden]
> sat in fund raising meetings with VCs in toronto

There's your problem. The single biggest thing every Canadian VC is trying to figure out is "why are these people asking us for money when if they were any good they'd be in the US" so by simply asking them you're already signalling something bad. A lot of their enthusiasm for process is based on this suspicion and also that the entire industry is just a way for various professional services to extract most of the investment money, since that's the game they're so used to playing with the government.

There are some Canadian VCs earnestly trying to improve but they are overwhelmingly hilariously conservative and focused on unimportant signals over reality. This is one (but not all) of the major factors that drive basically every remotely ambitious Canadian company to run a corp in Delaware and go for funding from the US. The tax situation is the other major contributor.

redanddead[hidden]
Canada doesn’t have throwing around money like in the US. We have resource extraction -> export money that’s it
cmrdporcupine[hidden]
There are boatloads of tax breaks and incentives that mitigate all the financial stuff and make running a startup here just fine honestly.

But that does nothing to make up for the terrible investment community. Getting started here requires already being started.

When I briefly worked for a Toronto startup, it was like all of them went to the same private boy's schools together as kids. It was a status club.

I jumped ship to an American startup and made almost double the money dealt with 0% of the bullshit and they were bought by Google the next year.

havercosine[2 comments hidden]
Its not normal time in SF/Bay Area. For better or worse, VCs in this city/region are thinking very differently on AI bets.

I think the key differentiator was that a team found a whitespace in what ChatGPT was doing, main comes from the same pedigree and team is as conscious of marketing as their product. SF VCs love these out of the box challengers, and people are claiming to replicate doesn't seem to matter.

The amount raised feels surprising but again entire SF/US AI scene is primarily "add moar layers and GPU" one trick ponies at this point.

majormajor[hidden]
> Its not normal time in SF/Bay Area. For better or worse, VCs in this city/region are thinking very differently on AI bets.

"Throw a bunch of money at copycats in the trend of the day" is very normal in SF/SV for the last several decades, going back to the dotcom boom.

The definition of "a bunch" has changed but so many "crypto for X" or "recommendations for X" or "uber for X" or "social for X", etc, things raised amounts that seemed wildly divorced from their market position.

HarHarVeryFunny[hidden]
A very major part of Jev is the cost and speed. Yes, classification is/will be a commodity business, just like LLMs are, and similarly there is no moat only production cost and pricing.

Yes, anyone can wrap a decisions API around an LLM, but so what? If you want to compete then you need to compete on price, and it's not clear if OpenAI and/or Anthropic are able or willing to do that without building a custom architecture, and even then is a race to the bottom on pricing really what they want to pursue?

I'm not sure if OpenAI have announced pricing for their Decisions API, but they have said it's based on Luna which costs $0.10/M input, not even remotely competitive with Jev's $0.04/M input, which I'd expect has some headroom built into it.

Assuming that the architecture behind Jev is not just an LLM, and gives them some inherent efficiency/cost and speed advantage, then the question is whether OpenAI and Anthropic really want to duplicate this and have a race to the bottom on pricing for what may be a large part of the business automation market they are addressing. Is that what they want as their IPO pitch - we're selling potatoes, and think can grow them cheaper than Typesafe ?

binlog[hidden]
Because every VC knows that one of OpenAI/Anthropic/Nvidia/Microsoft/Google/Meta/SpaceXAI/AMD/Stripe... will acquire them within the next year for talent alone.
smrtinsert[hidden]
When you frame it as if we're in the Pets.com era of AI the continued gold rush makes sense.
Culonavirus[hidden]
Oracle needs to hurry up and get downgraded to junk status at which point it will raise shit and at which point it will finally send off the avalanche of normalization across this industry.

There needs to be cleansing with fire. Weeds need to die. Trees need their branches cut. The sooner the better.

A shitty ass random "ai" startup built entirely on hype and astroturfing should not be raising anywhere near this amount of money.

dvt[9 comments hidden]
Is Jev being astroturfed on HN? It certainly feels like it. It's a middling product with virtually no moat (but great marketing).
minimaxir[hidden]
There have been a lot of posts/comments claiming "Jev-like models" but that's more of an shorthand for decision models, not astroturfing.
denverllc[hidden]
Yes, it's astroturfed everywhere (like X and reddit).

https://www.youtube.com/watch?v=xNgQtzEl4lY

Jev is used as an example of a successful marketing launch where they worked with many X "creators" prior to its release, so that all the creators would repost to put it to the top of everyone's feed. Then, over the following days they'd repost so it maintained momentum.

See: doomers.ai, clickstrike, growth matrix, etc. They use coordinated engagement, paid influencer networks, customized messaging, etc.

Jev isn't a terrible product, but it's way overhyped.

softwaredoug[2 comments hidden]
Why is it a middling product? Most clones don’t approach its performance, and it solves a specific problem well in a way that was awkward and ignored by most frontier labs.
dvt[hidden]
Because Jev is basically a "generalized classifier" which.. doesn't quite make much sense. Training a classifier is pretty easy and has been done routinely for like two decades now. Classifiers also tend to be very localized; for example, I've worked on classifiers that would bucket web traffic into "potential buyer" or "potential seller"—but this was very specific to the use case (vehicles, in our case).

If I seriously needed a classifier, I would just train my own and it would run on an iPhone. Any CTO worth their salt would suggest the same, because it's not even remotely comparable to training a large language model (w.r.t. compute or training data required).

vanuatu[hidden]
distribution is a moat
incompressible[3 comments hidden]
Just found out from a previous comment: https://doomers.ai/work/typesafe-ai-case-study

Obscene marketing is all you need it seems.

dvt[2 comments hidden]
Cool, yeah my spidey sense was definitely going off. It’s a half-decent idea, but it does appear that marketing is all you need.

Kind of sad seeing so many “tech people” (including here on HN) falling for it hook, line, and sinker.

hall0ween[hidden]
I appreciate you putting words to the astro-turfing. two areas llms have helped me are in (i) giving me work and (ii) learning how people in my life repeat what they hear from “knowledgeable” people. And not saying the voices proselytizing llms aren’t knowledgeable but I bet there’s an agenda behind their words (ie “i have something powerful” and “give me money!)
dovin[2 comments hidden]
Jev does seem to have become the Kleenex of decision models. Is brand recognition worth $7.5B? There are lots of other decision models out there that perform at or near jev-level (laya, gliner 2.5 decide, even embedding gemma 2) that you can also run locally, and honestly I think this kind of model makes the most sense running locally as well. Maybe if TypeSafe can ship fast they can stay the default. Guess we'll find out.
pickle-wizard[hidden]
I just started experimenting with the decision models. I spun up Laya on a VM with a couple of vCPU and 6GB of RAM. I get the results in about half a second. No need for GPUs or tons of memory.

I am integrating it into the product I am building and to me it doesn't seem like there is much need to go with a SaaS for this since the requirements are so light. I just can run it in Cloud Run and get all of the scale I'll ever need, and I get to tell my customers their data never leaves my environment.

xxmarkuski[hidden]
Their job opening as Safety Javeroni was a good laugh too [0]

[0] https://jobs.ashbyhq.com/typesafe-ai/9a94651c-5d63-4e82-8854...

dj_io[hidden]
There is significant congregation of investment towards companies like this whereas many other with meaningful innovation are struggling to survive. The model is not a revolution like I must say how llm did at the face value. Yes, transformers were there for long but the use case etc. If you do cost comparison using regular model and setting it with prompt and structured output you can get result very close. This is what most open Jev and other are doing with tuning llm.
maherbeg[4 comments hidden]
Has anyone actually eval'd the other open source options against Jev on real world tasks rather than looking at benchmarks?

I see a lot of people parroting the quick open source alternatives as being better on the benchmarks, but it's such a new category that I'm not convinced we have solid benchmarks.

I'm hoping a company releases an internal eval benchmark for these options. I'm sure some of the open source ones are solid in some cases, but would love to see more reliable data.

santiago-pl[3 comments hidden]
I found this benchmark helpful: https://benchmarkheaven.com/jev-models
TN1ck[2 comments hidden]
I also benchmarked them here [1] for content moderation. Jev is better than any of them.

[1]: https://tn1ck.com/blog/jevdit

maherbeg[hidden]
This is awesome! Please submit updates when a new model comes out that people are saying are better than Jev!
jacobgold[hidden]
Whether they can compete on decision models or not, TypeSafe showed that a lot of the market had missed something important. With this much money, they have a lot more chances to discover other important things that are missing.
minimaxa[hidden]
No one finds it strange that every defector from Opensi gets half a billion dollars in funding and already has a third of Fortune 500 companies paying for their product in a few weeks with a product that's already been replicated 10 times very easily?

OK, I'm jealous... Lol

nico[hidden]
Hopefully some of the Typesafe hype spills over this way to Jeffy[0]

Open source[1]. It comes with 68 pre trained classifiers which run and train on CPU alone. They run locally, are faster than Jev/Laya/Decisions, and they can perform many different tasks; from labeling email, all the way to playing Doom[2]

[0] https://jeffyclassify.com/

[1] https://github.com/nicobrenner/jeffy

[2] https://playground.jeffyclassify.com/#doom

redoxate[9 comments hidden]
What edge do they have over the market to justify such evaluation
gavinray[hidden]
Upvotes and Twitter hype
babelfish[hidden]
100mm arr
simonw[6 comments hidden]
Probably their team. Ex-OpenAI people who have already proven they can ship and get buzz for what they're doing.
miguelacevedo[5 comments hidden]
Same thing was said about Character AI, Cohere, etc. Companies started by the authors of the "Attention is all you need" paper. Look how that went... the team gets thrown around a lot as if it's the magic bullet. There is no magic bullet.
simonw[4 comments hidden]
The ex-OpenAI people who started Anthropic are doing pretty great right now.

VC is a hits business. Just one hit pays for 9 that didn't work out.

denverllc[2 comments hidden]
Anthropic's Series A was $124 million; Typesafe's is 7x that. If Jev becomes a $2bn company it will be a failure to investors.
simonw[hidden]
Anthropic raised their Series A a year and a half before ChatGPT had been released, when LLMs were still unproven technology and most VCs weren't paying attention to the field (they were mostly still chasing crypto).
majormajor[hidden]
This is a dangerous way of looking at things in a hits business.

The difference in hit rate matters enormously and the existence of a hit tells you nothing about the denominator.

The size of the bet matters too - you could wipe out a hit with one bad pick if that bad pick is big enough!

If you don't have an estimate of the rate that you trust, you're just throwing money at dreams.

But it's WAY harder to get rich by being a pessimist than by being an optimist. And if you're a VC choosing where to invest primarily-other people's money, then you have no particular reason to try to talk them out of the hype.

jghn[hidden]
Every time I see these headlines I wonder why the Scala company is back in the news
stickfigure[3 comments hidden]
Remember when reaching $1B valuation made you an exotic "unicorn"?
esafak[2 comments hidden]
The needle has moved. This is a wonderful thing.
shimman[hidden]
With any luck it'll radicalize a new generation of workers to rise up and stop SV.
atleastoptimal[hidden]
I think this is a hedge against major AI regulation.

Invariably near-AGI systems created by OpenAI/Anthropic will be very destabalizing. In the end the world will probably regulate AI capable of [any] <-> [any] input/output types. Models will need to be limited on their outputs by law so they cannot have unbounded, unpredictable outcomes. Jev is the ideal version of "benefits of AI without making humans obsolete" that might be the consensus once the track superhuman AI and its consequences are clear.

_davide_[hidden]
I guess they are investing in the team rather than on their demo product
woadwarrior01[hidden]
I wonder how much of it has been earmarked for astroturfing on HN and X? :D
jypepin[6 comments hidden]
Their headline says "TypeSafe A raises series AI". Is that AI slope?
CompleteSkeptic[3 comments hidden]
intentional to make it fun (:
stephen_cagle[hidden]
Is it a pun or something? I have been called dense.

I actually resized my browser thinking maybe something weird was going on with flex-wrap or overflow or whatever it is.

aidos[hidden]
With everything you must have going on right now, I love that you’ve come on to explain your headline is an intentional joke. All sorts being thrown around in this thread, but this was definitely the important thing to clarify.

For what it’s worth, however it works out, my guess is that the primitive Jev provides is likely to be considered essential in the future development of software.

throw03172019[2 comments hidden]
They also say they are a fun team. Maybe it’s a pun.
ambicapter[hidden]
Jokes all around: https://jobs.ashbyhq.com/typesafe-ai/9a94651c-5d63-4e82-8854...
2001zhaozhao[hidden]
Tbf this is THE way to get cost-effective AI into video games (think enemy AIs and NPC AIs that are much better than current ones and runnable on consumer GPUs), i'm really looking forward for decision models maturing.
rokhayakebe[2 comments hidden]
Can someone who actually knows these things share how might a company like this spend $870M over the years?
fnoef[hidden]
Koenigsegg for every founding member?
softwaredoug[hidden]
What’s interesting about the Jev moment isn’t just Jev, it’s the unleashing of distillation / fine tuning outside the frontier-adjacent labs. It’s the sudden explosion of a million Jevs.

If being an “AI Researcher” is a ticket to multimillion dollar salary, AI training talent cannot be contained to a handful of companies. It’ll become more common and diffuse. The old advice of not fine tuning, because it’s hard, goes out the window as that knowledge diffuses through the industry.

A similar thing is happening in search. For a long time labs have trained tailored embedding models. And now companies like SID training their own agentic models that are smaller and faster at search than GPT-5.

amelius[4 comments hidden]
They already got Sherlocked by OpenAI:

https://developers.openai.com/api/docs/guides/decisions

benatkin[hidden]
They weren't running on OpenAI, so nope.
vilos1611[2 comments hidden]
We've tested the decisions API against Jev at work and it's worse in various dimensions. Costs more, higher error rates, slower, and the answers are worse.

A lot of people are shouting about how Jev hasn't actually differentiated itself, but I question how much folks are actually experimenting with what's out there before coming up with an opinion.

For us, it's cleae that OpenAI rushed this out to meet the hype in the market right now without having a product that actually meets the bar Jev has set.

denverllc[hidden]
> but I question how much folks are actually experimenting with what's out there

I did. Originally I had a project that I had been wanting to do and thought to use a decision model for it. Jev, OpenAI, etc. are all within percentage points of each other.

Then I used traditional ML and found a small classifier (gemma 4) with traditional embeddings worked 2x as well.

Jev is the general purpose ML pipeline for when you want average results. Nearly every application has a "better" option available with a small amount of work.

greesil[hidden]
The amount of investment being thrown at this when we have serious existential problems as a species is how I know the system is definitively broken. Capitalism got us this far, but it's time for something new.
dmix[hidden]
That press release made me cringe a bit. Maybe corporate speak wasn't so bad after all.
maxdo[2 comments hidden]
interestingly , it destroy the landscape of Chinese models.

Unless china takes leadership in frontier space the picture is next :

1. cheap workhorses for classification, routing, other scenarios : Jev 2. coding agents with less erros : Anthropic/Openai, etc. 3. Science /Legal/Medical : A mixture of Jev+Anthropic scenarios

kylehotchkiss[hidden]
See clef
collimarco[hidden]
This looks like the top of the dot-com bubble...
bel8[hidden]
Do they have patents over jev related tech or something valuable to justify this?
alpineman[hidden]
From Andreessen the Trump donor. Sad, will switch to something else now.
enahs-sf[hidden]
got the recruiter call only to essentially be summarily rejected because my pedigree is wack. looks like i would've gotten hosed on valuation anyways.
jdw64[hidden]
I only use Jev at the tool-calling layer to keep tool costs low, but I don't really see what the big advantage is.
kevinkatzke[hidden]
A few year ago you could IPO at this valuation.
mococa[hidden]
~7B for a thing that we already have open-source?
robotswantdata[hidden]
I’ll just use open source, thanks
ada1981[hidden]
I built JEV-search and JEV-chat and have been really happy with the results…

Curious to see how all of these can work together, and giving LLMs the ability to deploy their own workflows and built type 1 systems as they go has also been fun and useful.

axus[hidden]
So... are they worth more, or less than 0xide
TeMPOraL[hidden]
Only just now I realized that "TypeSafe AI" are the people behind Jev, and "System One" isn't the company name, as I assumed, but a larger project label.
sajithdilshan[hidden]
Is there a second wave of AI bubble happening? How can an AI classifier company be worth of $7.5B
villgax[hidden]
Lol
gdiamos[hidden]
Good job typesafe.

You won the competition with VCs

Thorentis[hidden]
Another data point confirming that AI is a bubble.
pseudotensor[4 comments hidden]
Disclosure: I work at H2O.ai.

We released an Apache-2.0, open-weight 4B decision model that scores above Jev 1.13 on JevBench's composite score (72.5 vs 71.5) and is currently the top open model there: https://benchmarkheaven.com/jev-models . Newer models coming even larger than beat Jev in intelligence as well.

- Same contract as Jev: state + typed questions in, calibrated probabilities out, one forward pass, no generated tokens. - Your data never leaves your environment, and there's no per-call fee.

Weights, card and run instructions: https://huggingface.co/h2oai/h2o-lightning-4b

pseudotensor[hidden]
Update: Microsoft just announced Microsoft-Decision-1 (a post-trained Qwen3.5-9B) and benchmarked it against H2O-Lightning-4B, among others: https://commandline.microsoft.com/microsoft-decision-1-model...

One note on their speed comparison: the JevBench board shows self-hosted models with an adjusted latency of "2x + 0.15 s (assumption, not measured)". Our measured p50 is 29 ms; the adjusted figure is 0.21 s. They report 85 ms for theirs.

smrtinsert[2 comments hidden]
Is it possible to quantize the model? Don't think I want to run 8 gigs on my local just for decisions but curious about it for sure
pseudotensor[hidden]
Will do!