Company
San Francisco AI company, founded in 2024, that builds decision models intended to be embedded directly in software applications.
Workday's Joel Hellermark says its agents pair reasoning models with deterministic guardrails, itwire reported. Sana, bought for US$1.1 billion, now orchestrates every Workday agent, so the model-driven half of that split runs in one acquired product.
Reality
Workday said at Workday Rising that Sana, its agentic experience, is becoming the new Workday Home, with Sana Concierge to resolve HR and IT requests. Employees will work with Sana directly. That puts an agent ahead of the HR or IT team on routine requests.
Reality
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
Andreessen Horowitz led a financing of about $870 million in TypeSafe at a $7.5 billion valuation, Bloomberg reported. The money backs Jev, a model that hands software a choice plus a probability, and its use to builders depends on those probabilities matching how often it is right.
Perspective Coverage
8 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
OpenAI launched its Decisions API on October 6, three weeks after startup TypeSafe's cheap classification model Jev went viral with developers. Both chase the routine judgments a business automates by the thousand, where cost per decision decides what is worth doing.
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
Diogo Almeida left OpenAI two years ago convinced that human language is the wrong output for automation. The model his startup shipped this week returns probabilities, and one team testing it clocked classification 5 to 18 times faster.
Reality
OpenAI's Ultrafast tier, unveiled at DevDay 2026, sells faster GPT-6 Astra tokens at six times the standard API rate. Speed is now a separate line on the bill, so teams building coding pipelines have to decide which jobs are worth it.
Reality
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
Red Hat's 200M-parameter classifier scored 89.01% on prompt injection to a 35B LLM judge's 89.31%, answering in a median 54.1 ms against 312.5 ms. For injection screening, a large judge in the request path now has to justify nearly six times the latency for 0.3 points.
Reality
Diogo Almeida's TypeSafe put Jev into early access on 15 September with $40m from DCVC, and it sells a calibrated confidence number on every answer as the thing that makes automation possible, with the evaluations behind that claim built in-house.
Reality
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
Plain RAG and GraphRAG got none of 31 counting and superlative questions right on a 100-question TigerGraph hackathon benchmark, an entrant reports. A COUNT from the graph, set beside the evidence actually read, shows when an answer is incomplete.
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
NobodyWho rebuilt the core of TypeSafe AI's Jev with a local 0.6B Qwen model and 25 lines of Python. Teams pricing Jev for routing or tool-call safety checks can test that local baseline first, provided they measure its calibration on their own labeled decisions.
Reality
Sophos cites an independent test putting TypeSafe's Jev at 83% on Banking77 with no training examples, 10 points behind a trained classifier. Jev costs about a twelfth as much per email as Claude Haiku 4.5, so SOCs will be tempted to automate at a volume where small error rates add up.
Reality
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
TypeSafe AI is in talks to raise over $1 billion at a valuation above $10 billion, 50 times its seed price from two weeks earlier. Buyers at that price are paying for launch-week usage and an investor's profit claim, with no revenue figure published.
Reality
Build · September 22, 2026 · Confirmed14 publishers
Science · September 19, 2026 · Confirmed3 publishers
Build · September 21, 2026 · Confirmed2 publishers
Invest · September 21, 2026 · Confirmed3 publishers