A new class of model takes your application state and a structured question and returns a typed answer — a choice, a score or a yes/no — with a calibrated probability, in one fast pass.
Returns a valid answer and a confidence, so code can act on it or escalate to a person.
Writes text token by token — great for reasoning and writing, heavier for simple decisions.
Three inputs and outputs, whatever model sits underneath.
Pass the text or data the decision depends on, such as a support ticket, an email or a record.
Define the question and the allowed answers: an enum, a yes/no or a rating scale.
Receive the chosen answer plus calibrated confidence for each option, in a single pass.
one of N labelsClassify or route: pick the right category, queue or action.
probability 0–1A boolean with a probability, for example “does the user want a refund?”.
ordinal levelRate on a scale such as urgency, sentiment or quality.
The model the open-source projects below set out to match.
TypeSafe AI's first public System One model, built for “decisions inside software” rather than text generation. It returns typed decisions with calibrated probabilities so automation can tell when to act and when to hand off.
Two browser playgrounds, plus a Cloudflare-hosted decision model you can call through an API, so you can run typed decisions without installing anything.
A single playground and API for comparing decision models. Its site lists Jev, Laya, Kev, Solar and Span, with a workbench where you define typed questions and preview the request, and an example library that includes support triage, content moderation and lead scoring.
The site says the online playground is free to use. Check its pricing page for API access.
A playground, API guide and examples for Jev, the TypeSafe AI decision model. It offers a free in-browser demo, use cases, integrations and an agent skill.
This site describes itself as independent, so it is not TypeSafe AI's own site. The official Jev site is typesafe.ai, linked above.
by Cloudflare
A 27B multimodal decision model that turns a state and a schema of typed questions into decisions. It reads the state as text, JSON, images or video, and returns a probability for every allowed option of every question.
Best forTyped decisions over text, JSON, images or video on Cloudflare's platform.
Hosted on Cloudflare, so it is a managed alternative to Jev rather than an open-source one. Check the model page for current pricing and licence terms.
Three design families: small encoders that score options directly, LLMs read through their logits, and LLMs with dedicated decision heads.
by NandhaKishorM (Convai Innovations)
A non-autoregressive decision engine that scores the question and its options together in one forward pass. Supports choice, score and yes/no questions across 100+ languages, with calibrated probabilities and optional abstention.
Best forFast multilingual decisions on small models.
Ships an HTTP server compatible with Jev's, plus LangChain and MCP integrations. Its package page reports about 7.8× lower latency than Jev on its own benchmarks and notes Jev does better on label sets of 50+ options.
by Bespoke Labs
Adds Jev-style behaviour to an existing open LLM: it reads the model's scores for the permitted answer tokens and turns them into probabilities, instead of generating text.
Best forTraining, research and studying the data-curation recipe.
Publishes the model, training recipe, evaluation setup and data-curation pipeline.
by Jared Palmer
A family of small decision models you can train and run yourself. Handles yes/no, multiple-choice and rating questions in one request, and its server is designed so the TypeSafe Python SDK works against it unchanged.
Best forJev-compatible local deployment and fine-tuning on your own data.
Runs on CUDA, ROCm or Apple Silicon (MLX); the largest model handles up to 65,536 tokens of state.
by Theo Lee (independent)
A research experiment that needs no fine-tuning: it reads the logits for the allowed answers straight from an open model. The demo compares that direct readout with asking the model to write probabilities as JSON.
Best forExperiments and trying System One ideas on models you already have.
Formerly called OpenJev; independent and not affiliated with TypeSafe. The demo runs fully in your browser via wllama.
by Rizzo AI Academy
The open, local take on Jev: it limits each decision to a set of answer tokens, reads their logits and converts them to probabilities, without generating any text.
Best forLocal deployment on varied hardware, with a Jev-like HTTP interface.
Change the base URL to point existing Jev clients at a local server.
by wfzyx
A compact, non-autoregressive System One model that uses option markers to produce scores for choice, yes/no and score decisions directly.
Best forSmall, fast inference on CPU or modest GPUs.
Targets sub-15 ms decisions; runs on CUDA, ROCm, Apple Silicon and CPU.
by Tianyu (TianyuCodings)
A small, trainable Jev-style replica with parallel decisions, dynamic candidates and an end-to-end training pipeline.
Best forLearning how Jev-style architectures are built and trained.
Designed as a nano-scale replica for experimentation.
by Strands Labs
A small, fast decision model for agentic workflows. It picks between options or rates something on a scale, returning a calibrated confidence with every decision, and ships a command-line tool and an inference server.
Best forDecisions inside agents built with the Strands Agents SDK.
The project reports that answers at a confidence of 0.9 or higher were right about 95% of the time on short classification tasks it had not seen; treat that as the authors' own evaluation.
Licences read from each project's GitHub repository on 3 October 2026.
| Project | Approach | Base model | Licence |
|---|---|---|---|
| Laya | Encoder + decision head | ModernBERT-large (English), mmBERT-base (multilingual) | Apache-2.0 |
| Nimble | LLM logit readout | Qwen3.5-9B with rank-16 LoRA | No license listed on GitHub |
| Kev | LLM + decision heads | Qwen3.5 (0.8B, 4B, 9B) and Qwen3.8-27B | Apache-2.0 |
| SemIf | LLM logit readout | Qwen3 0.6B, MiniCPM5 2B or Qwen3.5 4B | MIT |
| Rizzo Flow | LLM logit readout | Open Spark-X2.5 models served with llama.cpp | Apache-2.0 |
| Von | Encoder + decision head | ModernBERT (about 395M parameters) | Apache-2.0 |
| NanoJev | LLM + decision heads | Qwen3-0.6B | MIT |
| Strands Decider | LLM + decision heads | Qwen3.5-2B-Base with pointer head | Apache-2.0 |
A decision model, also called a System One model, takes application state and a structured question and returns a typed decision — such as a choice, a score or a yes/no — together with a calibrated probability. It answers in a single fast pass instead of generating text token by token.
Jev is TypeSafe AI's first public System One model, built for decisions inside software rather than text generation. TypeSafe describes it as returning typed decisions with calibrated probabilities so software can decide when to act automatically and when to escalate to a person.
An LLM writes free text that your code must parse and may get wrong. A decision model is built to return only valid, typed answers with confidence, which makes it faster, cheaper and easier to wire into automation. LLMs remain better for open-ended writing and complex reasoning.
Choose Laya or Von for small, fast encoder models; Kev for a Jev-compatible local server with fine-tuning; Nimble to study a full training recipe; SemIf to experiment without training; Rizzo Flow for local llama.cpp deployment; and NanoJev to learn the architecture.
Not necessarily. Accuracy and speed figures are reported by each project on its own benchmarks, and the field is new. Laya's page itself notes Jev does better on label sets of 50 or more options. Test candidates on your own data before relying on one.
Several projects aim to be drop-in replacements: Kev says the TypeSafe Python SDK works against a Kev server unchanged, Laya ships a Jev-compatible HTTP server and Rizzo Flow mirrors the HTTP interface. Check each project's documentation for the exact API it supports.
Licences were read from each project's GitHub repository on 3 October 2026. Nimble's repository does not list a licence, so check with its authors before using it commercially.