↓ Skip to main content
9 – 17 UHR +49 8031 3508270 LUITPOLDSTR. 9, 83022 ROSENHEIM
DE / EN
DECISION TypeSafe AI USA

Jev (TypeSafe AI)

Jev by TypeSafe AI is not an LLM but a System One model: typed answers with probabilities, USD 0.042 per 1M input tokens, operated in the US, no EU region documented.

License Proprietary
GDPR Hosting Available
Context 64000 Tokens
Modality Text, JSON, Images (experimental) → Typed decisions (choice, score, noul) with probabilities, Bounding boxes (box, experimental)

Versions

Overview of available model variants

ModelReleaseEUStrengthsWeaknessesStatus
Jev 1.14 (jev-1.14.0, aliases jev-latest and jev-preview)
14 October 2026
Typed output instead of text: choice (up to 255 options), score (2 to 10 levels) and noul (yes probability 0 to 1), always schema-conformant Per TypeSafe, small improvements across the board over Jev 1.13, with the largest gain in multilingual support Calibrated probabilities per option, trained with Reinforcement Learning for Calibrated Decisions (RLCD) All questions of a call in one pass; per TypeSafe 70 to 500 ms end to end and 40 to 200 times faster than frontier LLMs on comparable tasks USD 0.042 per 1 million input tokens, output free (per the models overview in the docs) Experimental via headers: images as input (PNG, JPG), box question type for bounding boxes (Jev 1.14 only) and long context up to 262,144 tokens Near-frontier language understanding per the vendor; customisation via state, instructions and criteria in the request instead of fine-tuning Python and JavaScript SDKs, agent skill for Claude Code and Codex, versioned model field in every response for audits No training on customer requests per the docs, zero data retention for enterprise customers
No text output: not suited to drafting, summarising, code or multi-step planning Default context of 64,000 tokens per request (32,000 for state plus the longest question); long context and images not yet production-ready per TypeSafe, long context slower and less accurate English as the primary language; other languages including CJK with variable accuracy per the docs Closed weights, no self-hosting, no EU region: third-country transfer to the US Early access; rate limits set per organisation, no public default values documented Architecture and model size not published; calibration holds statistically, not per answer
Preview
Older, deprecated and retired versions (1)

Legacy and deprecated: superseded by newer models; for many versions the provider has already announced a shutdown. Retired: no longer available. Dates and successors are in the model lifecycle section.

ModelReleaseEUStrengthsWeaknessesStatus
Jev 1.13 (jev-1.13.0)
15 September 2026 (early access)
First released Jev version, same price and context limits as Jev 1.14 Can be pinned by version ID, for example for applications with confidence thresholds tuned to this version Images and long context also usable experimentally
Superseded by Jev 1.14 as the target of the jev-latest and jev-preview aliases No box question type Known weaknesses per the docs: counting, date comparisons, numeric values, a lean towards the first choice option
Legacy

Use Cases

Typical applications for this model

Ticket and email routing with confidence
Classification and prioritisation in workflows
Guardrail, judge and moderation decisions
Scoring on rubrics (quality, risk, urgency)
Decision nodes in AI agents
Feature extraction from large text volumes

Technical Details

API, features and capabilities

API & Availability
Availability Early access; REST endpoint POST https://api.typesafe.ai/v1/systemone, API key via the TypeSafe console, Python and JavaScript SDKs; experimental features via headers (TypeSafe-Danger-Zone-Images, TypeSafe-Danger-Zone-Long-Context)
Latency (TTFT) 70 to 500 ms end to end (vendor figure)
Features & Capabilities
Structured Output Vision
Training & Knowledge
Knowledge Cutoff not documented
Fine-Tuning Not available
Language Support
Best Quality English
Supported Other languages including CJK supported per the docs, accuracy varies
English is the primary language per TypeSafe; no vendor figures exist for German input, so run your own evaluation

Hosting & Compliance

GDPR-compliant hosting options and licensing

GDPR-Compliant Hosting Options
TypeSafe AI (API)
USA (operated from the US West Coast per the vendor)
Third-country transfer; DPA with EU standard contractual clauses, no training on customer data per the docs, zero data retention for enterprise customers; no EU region documented
License & Hosting
License Proprietary (API only)
Security Filters None (decision model without text generation)
Enterprise Support Yes
Cloud Only

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:

TypeAnswersReturnsLimit
ChoiceWhich of these options applies?chosen option, probability per option, confidenceup to 255 options
ScoreWhich level on an ordered scale?value, probability per level, confidence2 to 10 levels
NoulIs 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 heightone 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

VersionModel IDReleasedStatus
Jev 1.14jev-1.14.014 October 2026current, target of jev-latest and jev-preview
Jev 1.13jev-1.13.015 September 2026still 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.

Cost estimation for this model

For up-to-date token pricing, model variants and EU availability, see our sister project ai-prices.eu. It helps you compare and estimate the operational cost of leading AI models for your specific use case.

Compare prices on ai-prices.eu

ai-prices.eu is a project by innFactory AI Consulting GmbH and provides transparent cost estimates for leading AI models.

Frequently Asked Questions

What is Jev?

Jev is the first System One model from TypeSafe AI, in early access since 15 September 2026; the current version is Jev 1.14 (jev-1.14.0) of 14 October 2026. It takes unstructured text or JSON as state and answers predefined questions with typed values and probabilities: one option from a list (choice), a level on a scale (score) or a yes probability (noul). Experimentally, Jev also processes images and, with the box question type, returns the position of an object in an image. Jev writes no text and no code.

What does Jev cost?

Per the TypeSafe documentation, Jev 1.14 and Jev 1.13 each cost USD 0.042 per 1 million input tokens (USD 42 per 1 billion); output tokens are not charged. The optional long-context mode costs the same per token. TypeSafe sets rate limits per organisation and shows them in the developer console, with higher limits available on custom and enterprise plans.

Can Jev be used in a GDPR-compliant way?

That needs a case-by-case review. Jev is only available as an API from TypeSafe AI; no EU region is documented and, per the vendor, the service is operated from the US West Coast. TypeSafe provides a data processing agreement with EU standard contractual clauses, states that it does not train on customer requests and offers zero data retention to enterprise customers. Anyone who must keep data in the EU can self-host the open model Laya instead.

How does Jev differ from an LLM?

An LLM generates text token by token that software first has to parse and validate. Jev returns only values from a predefined answer space and adds the probability distribution. All questions in a call are answered in parallel in a single pass, and the output always fits the schema. Per TypeSafe it is trained with Reinforcement Learning for Calibrated Decisions (RLCD) on calibrated decisions rather than human preferences.

What are Jev's limits?

Jev is not suited to tasks that need text as a result: drafting answers, summarising, writing code, multi-step planning. The default context is capped at 64,000 tokens per request (32,000 for the state plus the longest question); a header extends it to 262,144 tokens, which per TypeSafe is slower and less accurate. A choice question is capped at 255 options, a score at 2 to 10 levels. Per TypeSafe, images and the box question type are not yet production-ready. English is the primary language; other languages including CJK are supported per the docs with variable accuracy. The model can still decide wrongly; calibration holds across many predictions, not for a single answer.

Consultation for this model?

We help you select and integrate the right AI model for your use case.