Topic
Non-generative models that score or classify options and output probabilities, serving as a judgment layer alongside generative language models.
Subquadratic's SubQ Snap decision model accepts 384,000 tokens per request, 12 times Jev's 32K limit, according to the company. CEO Alex Whedon says the cap reflects preview compute capacity. The accuracy and latency figures are Subquadratic's own.
Reality
Workday gave TypeSafe's Jev every atomic component in its design system and lets the model pick what Sana shows, said chief AI officer Joel Hellermark. TypeSafe claims 70 to 500 milliseconds per answer; whether Workday sees that depends on how many screens reduce to closed questions.
Reality
OpenAI, Microsoft and Perplexity launched decision-model APIs within nine days at $0.10, $0.042 and $0.02 per million input tokens. Perplexity's 27B weights are Apache 2.0, so a buyer with a roughly 49 GiB GPU can self-host v1.1.
Reality
Subquadratic launched SubQ Snap at $0.05 per million input tokens with a 384K-token window, the sixth hosted decision model in about three weeks. No outside lab has run it, so every score in the launch is a claim about Subquadratic's own test set.
Reality
Microsoft has put Decision-1, a model that scores fixed choices, in Foundry at $0.042 per million input tokens. Agent teams now have to test whether its scores, so far benchmarked only by Microsoft, are calibrated well enough to decide when a case goes to a human.
Perspective Coverage
7 publishersReality
Microsoft released Microsoft-Decision-1, a small model that returns calibrated probabilities for fixed options, priced at $0.042 per million input tokens. The accuracy and speed figures are Microsoft's own, and the base model is due to change soon.
Reality
Cloudflare's Clef-omni takes audio and video alongside text and images and returns schema-bound decisions in one API call. Runtimewire reported that the same release cut hosted Clef-flash's context window from 64,000 to 24,000 tokens, so longer prompts now go to Clef or to self-hosted weights.
Perspective Coverage
3 publishersReality
Liquid AI's open-weight d1-3B answers classification questions in a single forward pass, taking 16 ms on a Jetson AGX Thor and 50 ms on an Orin Nano. That puts routing and intent tagging on edge boards without a large model, on figures Liquid measured itself.
Perspective Coverage
3 publishersReality
AWS, Upstage and Ollama now serve decision models through TypeSafe's System One request format, within three weeks of TypeSafe launching Jev. OpenAI has not signed on, so teams wiring decision calls into agents are picking an interface while the standard is still forming.
Reality
Cloudflare released Clef, open-weight 9B and 27B models that return a probability for each allowed option in a single forward pass. An agent's routing or escalation call becomes a score it can threshold, and so far the only latency figures come from Cloudflare.
Reality
OpenAI's Decisions API answers yes/no, pick-one and scale questions for $0.10 per million input tokens, with output free. Jev sells the same three question types for text at less than half that, so OpenAI's case rests on image input and compliance terms.
Perspective Coverage
4 publishersReality
Ollama and llama.cpp added local endpoints within a week for Jev decision models, 144M-to-27B checkpoints that return a label and a probability. Routing calls can move off hosted models once teams check each checkpoint's licence and test it on their own labels.
Reality
Musubi released PolicyLM-1.7B under Apache-2.0 to score messages against a written policy in a reported 22 milliseconds on an H100. Teams can rewrite rules without retraining, though every speed and accuracy figure so far comes from Musubi's own tests.
Perspective Coverage
3 publishersReality
Cloudflare released two Jev-API-compatible decision models, Clef and Clef-flash, on Workers AI and as Apache 2.0 weights on Hugging Face. Typed classification steps in agent code can now move between providers or onto owned hardware, as long as they stay inside the text-only, 32k-context features Jev supports.
Perspective Coverage
4 publishersReality
OpenAI and AWS shipped their own decision models within two weeks of TypeSafe AI's September 15 launch of Jev, according to a dev.to account. For routing work, the comparison with an LLM call turns on whether a decision model's probabilities are calibrated.
Reality
AWS open-sourced Strands Decider 2B, a model that picks from a fixed list of options in under 150 milliseconds on a local machine, the company says. Agent builders get a small model they can run themselves for routing and tool-selection steps that would otherwise each call a full LLM.
Perspective Coverage
3 publishersReality
llama.cpp merged a /v1/systemone endpoint on October 2 that returns typed answers with probabilities from five open model families running locally. Whether it can replace hosted classification calls depends on calibration that each team has to measure on its own data.
Perspective Coverage
3 publishersReality
TypeSafe AI's Jev, a model that returns scored decisions, is in some developers' products after a launch video drew almost 40 million views on X. The evidence so far is hackathon testimony, so the useful work for leaders this quarter is finding which routine decisions can be split into narrow questions.
Reality
Mert Cobanov's open-source Ollaya serves local decision models behind a Jev-compatible API and reports 8 to 10 ms per request on an RTX 4090. Its comparisons with TypeSafe's hosted service are uncontrolled, so teams have to check accuracy on their own traffic before switching.
Reality
TypeSafe's Jev answers typed questions with floats and probabilities. An invoice pipeline that handed it classification and catalogue selection still needs a generative model for field extraction and for the note a human reads.
Perspective Coverage
14 publishersReality