The System One models ecosystem.
Loading…
The System One models ecosystem.
Featured models
Picked by the System One Models team. Accuracy is as each publisher reports it, on its own test suite, so check the suite before comparing numbers.
1. 2
julia-1Supersonic Labschoice · score · noulOpen73.2%
2. 3
jevTypeSafe AIchoice · score · noulAPI—
3. 4
tev1Together AIchoice · classify · routeOpen + API88.0%
4. 5
layaConvai Innovationschoice · score · noul
A decision, not a paragraph.
A System One model is a decision model: it reads a state and answers typed questions — choose an option, score on a scale, or give the probability that a condition holds — in one forward pass, with calibrated confidence and no generated text.
choicepick one option from a list you send
scoreplace the state on a scale you define
noulthe probability that a condition holds
› Which way should the snake go next?
The snake is heading east, and the food looks to be above it and slightly to the left. Turning left would bring it closer, but it is worth checking first whether its own tail could block the way back. On balance, the most reasonable choice is probably to turn left — although carrying straight on for one more step would also be defensible.
left0.06
straight0.06
right0.89
`{"action": "right", "confidence": 0.89}`
decision #11.3 ms
Illustration of each model’s output. Latencies vary by model and hardware. Why they need their own home
From finding one to publishing your own.
One site and one CLI for the whole workflow.
1. 1
Find
Every System One model with what it decides, its licence, size and price. Featured and trending lists, and filters by capability and architecture. Browse models 2. 2Compare
One table of every model, head-to-head pages, or any four side by side. A benchmark is always shown with the suite it was measured on. The comparison table 3. 3Try
A playground on every model page. Models on the System One Engine answer live once you sign in; for the rest it builds the request and the code to send it. Try anyjev 4. 4Pull
Every file checked against its SHA-256 and cached once, with versions pinned. `systemone pull namespace/name` 5. 5Fine-tune
System One Studio, open source, on your own machine: MLX on a Mac, PyTorch on NVIDIA, AMD, Intel or the CPU. LoRA, DoRA, rsLoRA and LoRA+, scored against the base before you publish. `systemone run studio` 6. 6Publish
To your own page or your organisation’s, as an immutable version. The CLI writes the manifest and model card from what it finds. `systemone push ./my-model`zsh — systemone
$ pip install systemonemodels Successfully installed systemonemodels-0.2.4 $ systemone pull convai-innovations/laya 19 files · downloaded 2.4 GB $ systemone run studio System One Studio http://127.0.0.1:8765 $ systemone login ✓ Signed in as you $ systemone push ./my-router ✓ Published you/[[email protected]](https://systemonemodels.tech/cdn-cgi/l/email-protection) with 4 files
$
The CLI guide · System One Studio
Who builds them
- Supersonic Labs1 model
Bespoke Labs1 model
Nokia1 model
iApp Technology- Sep 27, 2026 · 4 minWhat is an AI decision model?A model that returns an answer from a set you define, with a calibrated probability, instead of writing text. How it works, how it differs from an LLM, a classifier and JSON mode, and when to use one.
- Sep 27, 2026 · 5 minOpen-source Jev alternatives, and how to choose oneJev's weights are not published. Within ten days of its release, more than twenty open System One models appeared that you can run and fine-tune yourself, and eleven of them answer Jev's own API. What each one is, and how to pick.
- Sep 23, 2026 · 3 minWhy decision models need their own registryA general model hub can store a decision model perfectly well. Storage was never the problem — the metadata, the evaluation and the workflow are.
- 6
kevJared Palmerchoice · score · noulOpen83.8% - 7
decisionvLLM Semantic Routerchoice · score · noulOpen77.4% - 8
gliner2-5-decideFastino Labschoice · score · noulOpen + API60.2% 
Cua1 model- Fastino Labs1 model
Contrastive-LM1 model- Together AI1 model
Blockbrain Labs1 model
akhilaaa31 model
EldanRing1 model
Kotoba Labs1 model
From the blog
Questions developers ask first
What is System One Models?
System One Models (systemonemodels.tech) is the developer platform for System One models: a registry to find and compare them, a CLI to pull and publish them (`pip install systemonemodels`), hosted files with checksums, a playground on every model page, and System One Studio to fine-tune them on your own machine.
Is there a Hugging Face for System One models?
Yes: System One Models (systemonemodels.tech). Hugging Face hosts models of every kind; systemonemodels.tech is a registry and developer platform built only for System One models. It lists every one, open weights and hosted APIs such as Jev, with what each decides, its licence, price and reported benchmarks. It compares them side by side, pulls them with one command (`pip install systemonemodels`), and fine-tunes and publishes them with System One Studio. Open models link back to their Hugging Face repos.
What is a System One model?
A System One model is a decision model: it reads a state and answers typed questions — choose an option, score on a scale, or give the probability that a condition holds — in one forward pass, with calibrated confidence and no generated text.
How is a System One model different from an LLM?
A language model generates text token by token, so its answer has to be parsed and takes longer the more it writes. A System One model returns a typed value — an option, a score or a probability — in one pass, usually in milliseconds, with a confidence your code can put a threshold on.
Which System One models are there?
Jev from TypeSafe AI was the first, in September 2026. Open models and reproductions followed, among them Laya, Kev, GLiNER2.5-Decide, CLM, Bespoke Nimble and Decider. The live list, with prices and reported benchmarks, is at systemonemodels.tech/system-one-models.
How do I download a System One model?
`pip install systemonemodels`, then `systemone pull namespace/name`. Every file is checked against its SHA-256 and cached, and `--version` pins a release.
How do I publish or fine-tune one?
`systemone push ./my-model --repo you/name`
Build System One models?
Your organisation may already have a page here, with your models on it. Claim it, and publish your next release where people compare them.
Find your organisationCreate an account
Everything you need for System One models is here. Find them, compare them, try them in your browser, pull them with one command, fine-tune your own and publish it back.
$ pip install systemonemodels
Compare every modelOpen-source Jev alternativesWhat is an AI decision model?
Featured #1All featured
anyjev
by Nokia
A training-free layer that turns an open LLM into a decision model. It reads typed choice, yes/no and score answers from one prefill, removes option-order bias with no labels and, from a few hundred labels, calibrates or fits a closed-form head. Weights stay untouched.
choicescorenoulclassifyroute
Reported accuracy77.1%LocalLLaMA/typed-decisions (2,000 held-out decisions), L2 with 300 labels per question
Size8.0B
WeightsOpen weights
Licenceapache-2.0
Try itModel card 0 0
Open
—
1 model
publishes a folder; the CLI writes
`systemone.yaml`
and a model card from what it finds. To fine-tune Laya on your own data, System One Studio trains it locally and publishes the result here in one step.
Does it cost anything?
Browsing, downloading and publishing public models are free. Hosted models are billed by their own providers, at the prices listed on each model page.