Preview. These maps are our reading of public sources and haven't been checked with the projects' maintainers yet.
Kev
Family of small Jev-like decision models (0.8B to 27B, Qwen3.5/3.8 base with adapter and pointer head) you can run or fine-tune yourself; serves TypeSafe's System One API so the official SDK works unchanged.
Release kev-1.0 on 2026-10-01; repo pushed 2026-10-23; 9.5k stars. kev-4b on Hugging Face ~20k downloads. (checked 2026-10-25)
In The Jev ecosystem
Where it stands apart
- What is the main thing it ships? model weights (19% of 21 others)
- Does it train on Jev's outputs? (TypeSafe's customer agreement forbids using output to distill or build competing models.) it trains models, but not on Jev's outputs (21% of 19 others)
What it shares with most of the community
- Is it actively maintained? active (62% of 21 others)
- How is the main thing licensed? permissive open source (MIT/Apache/BSD) (86% of 21 others)
- How is it mainly sustained? individual side project, no revenue (55% of 20 others)
- Can it run entirely on your own device or infrastructure, without the maker's servers? yes (52% of 21 others)
- Do you need an account to use it? no account (86% of 21 others)
- Who is it for? developers (95% of 21 others)
- Which shared standard or format does it mainly build on? TypeSafe's System One API (state + typed questions, /v1/systemone) (81% of 21 others)
- Which models answer its questions? Pick the first that applies. other models, not Jev (open, local, or LLMs) (52% of 21 others)
- Which question types does it handle? Jev's three: noul, choice, score (67% of 21 others)
- Can you use it without a TypeSafe account? yes, no TypeSafe account involved (62% of 21 others)
- Are its answers typed values rather than generated text? (identity question) yes, typed answers only (67% of 21 others)
Nearest neighbors
- Gutsy 94% of 16 shared answers match
- Laya 94% of 16 shared answers match
- snap 81% of 16 shared answers match
Every answer
| Question | Answer | Source |
|---|---|---|
| Is it actively maintained? | active | source 2026-10-25 |
| How is the main thing licensed? | permissive open source (MIT/Apache/BSD) Apache-2.0 code and weights. | source 2026-10-25 |
| How is it mainly sustained? | individual side project, no revenue Personal repo by one author, Apache-2.0, no pricing. | source 2026-10-26 |
| Can it run entirely on your own device or infrastructure, without the maker's servers? | yes | source 2026-10-25 |
| Do you need an account to use it? | no account Runs locally; API key is optional and self-set. | source 2026-10-26 |
| Who is it for? | developers | source 2026-10-25 |
| Which shared standard or format does it mainly build on? | TypeSafe's System One API (state + typed questions, /v1/systemone) 'The API matches TypeSafe's System One.' | source 2026-10-26 |
| Which models answer its questions? Pick the first that applies. | other models, not Jev (open, local, or LLMs) | source 2026-10-25 |
| What is the main thing it ships? | stands apart model weights Plus serving and fine-tuning code. | source 2026-10-25 |
| Which question types does it handle? | Jev's three: noul, choice, score | source 2026-10-25 |
| Does it check whether its probabilities can be trusted? Pick the first that applies. | yes, it publishes calibration metrics (ECE, Brier, or similar) Ships a fitted temperature per checkpoint; reports Brier, ECE and confident-error rates vs Jev. | source 2026-10-25 |
| Where do the decisions run by default? | on your own hardware (CPU, Mac, or GPU) CUDA or Apple Silicon (MLX); optional one-command hosted endpoint on Modal. | source 2026-10-25 |
| Does it train on Jev's outputs? (TypeSafe's customer agreement forbids using output to distill or build competing models.) | stands apart it trains models, but not on Jev's outputs Trained on public datasets and generated policy examples. | source 2026-10-25 |
| Can you use it without a TypeSafe account? | yes, no TypeSafe account involved | source 2026-10-25 |
| Are its answers typed values rather than generated text? (identity question) | yes, typed answers only | source 2026-10-25 |
| Does it publish head-to-head numbers against Jev? | yes, head-to-head numbers against Jev Kev-27B 0.851 vs Jev 0.857 on new sources. | source 2026-10-25 |
Sources
This page was derived from:
- GitHub · github.com/jaredpalmer/kev · read 2026-10-25
RSS for this project · Its own release feed
Are you a maintainer? Claiming and correcting projects is coming soon.