# Jev Notes

> Unofficial notes for TypeSafe’s Jev AI — decisions, not chat.

Jev AI is TypeSafe’s System One decision model (Jev 1.13): Structured decision model: returns typed answers with probabilities; does not generate free-form text. Primitives: noul (yes / no questions), choice (up to 255 options; use null for options without detail), and score (ordered scale: at least 2 and at most 10 levels). Price: $0.042 per 1M input tokens; output tokens are free.
Unofficial developer resource. Not affiliated with TypeSafe AI.

Full text: https://jevnotes.com/llms-full.txt

## Guides

- [Jev AI by TypeSafe: Decision Model Notes](https://jevnotes.com/handbook/): Unofficial handbook for Jev AI and TypeSafe AI: System One decision model API, noul/choice/score primitives, pricing, and when typed decisions beat a chat LLM.
- [TypeSafe AI: The Company Behind Jev AI](https://jevnotes.com/typesafe-ai/): Who TypeSafe AI is, founder Diogo Almeida, System One naming, published speed/cost claims, and official docs links.
- [Jev API Quickstart: Python System One Example](https://jevnotes.com/api-quickstart/): POST /v1/systemone with noul, choice, and score in one call. Auth, errors, backoff, and why chat-completions SDKs will not work.
- [What Is Jev AI? Noul, Choice and Score](https://jevnotes.com/what-is-jev-ai/): Jev AI is not a chat model. Learn noul, choice, score, confidence vs probability, and when not to use it.
- [Jev Pricing Calculator: $42 per Billion Tokens](https://jevnotes.com/pricing-calculator/): Estimate daily and monthly cost at $0.042 per million input tokens. Output tokens are free. Compare against a chat LLM.
- [Jev Primitives: Noul vs Choice vs Score](https://jevnotes.com/primitives/): Field reference for noul, choice, and score plus a request builder that copies valid System One JSON.
- [Jev AI vs LLM: When Decision Models Win](https://jevnotes.com/jev-vs-llm/): Parallel sampling vs autoregression, latency, calibrated probabilities, and using LLM write + Jev judge together.
- [Jev as a Judge: Rubric Scoring and Evaluation](https://jevnotes.com/jev-as-a-judge/): Public posts in this catalog that mention a judge, a rubric, or an eval, plus the score primitive and the sel-jev-rubric-judge traffic label stored here.
- [Jev AI on OpenRouter: Decisions API Guide](https://jevnotes.com/jev-openrouter/): Recorded OpenRouter model ids for Jev AI, the Decisions API warning, and prices stored in this handbook. Re-check the live card before you ship.

## Cases

- [Jev AI Examples: 413 Real Builds & Demos (Updated Daily)](https://jevnotes.com/): 413 real Jev AI builds and public demos, updated daily. Original posts, metrics, and source links. Unofficial notes, not TypeSafe docs.
- [Training](https://jevnotes.com/cases/category/train/): 13 cases
- [Agents](https://jevnotes.com/cases/category/agent/): 81 cases
- [Guides](https://jevnotes.com/cases/category/guide/): 38 cases
- [Apps](https://jevnotes.com/cases/category/product/): 215 cases
- [Benchmarks](https://jevnotes.com/cases/category/benchmark/): 32 cases
- [Infrastructure](https://jevnotes.com/cases/category/platform/): 18 cases
- [Notes](https://jevnotes.com/cases/category/note/): 16 cases

## Projects

- [Jev GitHub Projects (2026): 256 Repos & Skills](https://jevnotes.com/projects/): Official TypeSafe SDKs (Python/JS), 256 listed repos, featured stars and last push synced from GitHub, and skills listed.
- [Jev AI GitHub Skills: 5 Repos (Agent Skills)](https://jevnotes.com/projects/skills/): 5 Jev skill repositories on GitHub, grouped by category, including official typesafe-ai/skills when the catalog lists it. Unofficial, not TypeSafe AI.

## Blog

- [Jev Notes Blog](https://jevnotes.com/blog/): Dated developer notes on Jev AI channels and API shapes.
- [Blog: jev-channels-september-2026](https://jevnotes.com/blog/jev-channels-september-2026/): Unofficial dated developer note.