innFactory AI Consulting from Rosenheim is tracking Jev because the model addresses a gap we see in automation projects every day: most decisions in a workflow do not need text, they need a value that software can act on. For a detailed explanation of how Jev differs from an LLM, read our blog post Jev by TypeSafe: the AI model that writes no text. This page summarises the facts from TypeSafe AI’s announcement, the Jev 1.14 post and the documentation. This overview reflects the state as of October 2026.
What is Jev?
Jev is the first System One model from TypeSafe AI, founded by Diogo Almeida, who by his own account worked at OpenAI on the methods that taught language models to follow instructions. The name “System One” refers to Daniel Kahneman’s distinction between fast, intuitive judgement (System 1) and slow, deliberate thinking (System 2): reasoning models are System 2, Jev is meant to be System 1.
The vendor’s short formula: a function call with frontier intelligence. Unstructured state in, typed probabilistic decisions out. Jev generates not a single token of text, so sequential generation disappears and all questions in a call are answered in parallel.
How a call works
A request to POST /v1/systemone contains a state (text, JSON object or text array), the model (jev-latest or a version ID such as jev-1.14.0) and a map of questions. There are three question types for decisions, plus the experimental box type:
| Type | Answers | Returns | Limit |
|---|---|---|---|
| Choice | Which of these options applies? | chosen option, probability per option, confidence | up to 255 options |
| Score | Which level on an ordered scale? | value, probability per level, confidence | 2 to 10 levels |
| Noul | Is this statement true? | probability of “yes” between 0 and 1 | |
| Box (experimental) | Where in the image is the object? | the four edges of a box as fractions of image width and height | one box per question, Jev 1.14 only |
The response carries the same keys as the questions, plus a model field with the version actually used (currently jev-1.14.0 for jev-latest) and usage with input and output tokens. Customisation happens through state, instructions and criteria in the request, not through fine-tuning. With the header TypeSafe-Danger-Zone-Images: true, images (PNG or JPG) can be placed anywhere text can go: in the state, the instructions and the criteria. A request may be at most 32 MiB. Error codes: 400 (request rules, such as an exceeded context length), 401 (key), 413 (body too large), 422 (validation), 429 (rate limit), 529 (overload).
Pricing, limits and availability
- Price: USD 0.042 per 1 million input tokens for Jev 1.14 and Jev 1.13; output tokens are not charged (per the announcement, “too cheap to meter”).
- Context: by default 64,000 tokens combined per request, of which at most 32,000 for the state plus the longest question. With the header
TypeSafe-Danger-Zone-Long-Context: true, both limits rise to 262,144 tokens at the same per-token price; per TypeSafe this is slower and less accurate. - Inputs: text, JSON objects and text arrays; images experimental, no audio or video.
- Rate limits: per organisation, shown in the developer console; higher limits on custom and enterprise plans.
- Latency: 70 to 500 ms end to end per the vendor; TypeSafe claims 40 to 200 times faster processing than frontier LLMs on comparable System One tasks and, in a workflow example, a factor of 193.6 (time) and 444.6 (cost). These are vendor figures without independent measurement.
- Access: early access since 15 September 2026 (with a waitlist per the announcement), API keys via the TypeSafe console; SDKs for Python and JavaScript, an agent skill for Claude Code, Codex and other agent environments.
- Languages: English as the primary language; other languages including CJK with variable accuracy per the docs. For Jev 1.14, TypeSafe expects the largest gain in multilingual support.
Versions and lifecycle
| Version | Model ID | Released | Status |
|---|---|---|---|
| Jev 1.14 | jev-1.14.0 | 14 October 2026 | current, target of jev-latest and jev-preview |
| Jev 1.13 | jev-1.13.0 | 15 September 2026 | still available, superseded |
Jev 1.14 first shipped under jev-preview; jev-latest now points to it as well. Per TypeSafe, the release brings small improvements across the board; the box question type works only with Jev 1.14. Jev 1.13 remains available for now. TypeSafe says it will give 60 days’ notice before formally deprecating jev-1.13.0 and 30 days for jev-1.14.0, and intends to publish a deprecation policy. Because answers can change when an alias moves, the docs recommend sending the version ID instead of the alias if you have tuned confidence thresholds against one version.
Data protection and operation
Jev is available exclusively as a hosted API. No EU region is documented; per TypeSafe the service is operated from the US West Coast. For the GDPR assessment that means a third-country transfer. TypeSafe provides a data processing agreement with EU standard contractual clauses, states in its docs that Jev is not trained on customer requests or responses, and offers enterprise customers zero data retention. Since Jev 1.14, customers can opt in to share requests for debugging; per TypeSafe, these are still not used for training. The public docs do not mention security certifications such as SOC 2. Before production use with personal data, the DPA, retention periods and purpose of processing belong in the data protection impact assessment.
Positioning: classifier, not LLM
Jev cannot invent a value outside the schema; in that sense there is no hallucination and no type error. The model can still be wrong: it then picks the wrong option. TypeSafe itself writes that calibration is measured across many predictions and is no guarantee for a single answer. The confidence it returns therefore belongs in every decision logic: automate high confidence, escalate low confidence to people.
Jev suits fast, narrowly scoped judgements inside software: classify, route, score, check, set guardrails, turn large data volumes into features. It does not suit anything that needs text as a result. That remains the job of LLMs such as GPT, Claude or Gemini; in many architectures a System One model sits as a router in front of the expensive reasoning model.
Open alternatives: Laya and Clef
If you want to use the System One idea but keep data in the EU, there are two open options with a Jev-compatible endpoint. Laya by Convai Innovations is a small encoder model (322 to 421 million parameters) under Apache 2.0 that usually needs fine-tuning for good results. Clef and Clef-flash by Cloudflare (since 1 October 2026, 27 and 9 billion parameters on Qwen bases, also Apache 2.0) additionally process images and video, run hosted on Workers AI or self-hosted, and per Cloudflare lead Jev on 7 of 10 decision benchmarks; Jev leads on knowledge benchmarks. Details and each vendor’s published comparison figures are on the model pages.
Our recommendation
Jev pays off where an LLM is currently misused as a classifier, judge or router and latency or cost hurt. Check three points before adopting it: whether the inputs are free of personal data or the contractual situation (DPA, ZDR) is sufficient for US processing, whether your language is covered well enough, and whether the answer space can really be fixed in advance as choice, score or noul. TypeSafe itself recommends images, box questions and long context for prototypes only, not for production. If you want to know which decisions in your workflows can be offloaded this way, get in touch.
Related decision models
The decision model (System One) category also includes Microsoft-Decision-1.1, Gemini Decision 1, Cloudflare Clef, Laya, Julia-1, GLiNER2.5-Decide. OpenAI is moving in the same direction with the Decisions API based on GPT-6 Luna; details are on the OpenAI GPT page.
