A team tested decision models Jev and Luna against the frontier model Opus on grading identity resolution results, a task that maps scattered data records to real-world identities. Used alone, Jev and Luna scored 88.8-97.3% accuracy, but a hybrid system letting Jev escalate uncertain cases to Opus matched Opus-level accuracy while cutting costs by more than 4x. The tests also found decision models run up to 90 times cheaper and 10 times faster than Opus, though they lag Opus by about 6 percentage points on harder cases.
hightouch.com
· 2026-11-06
OpenAI made its Decisions API generally available to all customers, adding voice and visual input options and reducing global latency through geographically targeted inference. The API repurposes ChatGPT Luna to deliver fast, low-cost classification-style decisions, cutting response times from seconds to under 200 milliseconds, according to OpenAI product lead Nikunj Handa. Early customers including Box, Datadog, Ramp and Speak have already used it to replace standard LLM calls in their workflows.
techcrunch.com
· 2026-10-22
Vercel reported that its Jev decision model, released in limited form on September 15 by TypeSafe AI, reached 35% of paid teams within two weeks — nearly double the reach of the next most popular model, Sonnet 5.5. The data, drawn from Vercel's AI gateway, also shows open source models now handle over 70% of tokens processed but only about 25% of total spend, while frontier-lab spending has grown tenfold.
techcrunch.com
· 2026-10-16
A developer ran the Trolley Problem thought experiment through Jev and twelve other 'decision models,' including OpenAI's Luna and eleven other local models, plus a baseline called Laya. These decision models are a new category of AI that output discrete choices with confidence probabilities rather than text, and are marketed as much faster and cheaper than large language models while retaining comparable reasoning ability. The experiment repeated the classic ethical dilemma in multiple phrasings across all models to compare their responses.
zachlandes.com
· 2026-10-15
The Jev team released version 1.14.0 alongside a set of experimental 'danger zone' features gated behind API header flags, including composable image support, a new data type, and longer context windows. The company says these features are not yet reliable enough for production use but are intended to evolve into stable capabilities over time.
typesafe.ai
· 2026-10-14
A workflow pairs Unsiloed, which extracts structured fields from documents and locates them on the page, with TypeSafe's Jev, which evaluates extracted data against targeted questions. Jev separates input data (state) from the checks it runs (questions), returning typed answers such as probability scores, multiple-choice selections, or rubric-based ratings. The combined output is used to flag which extracted fields likely need human review, illustrated with a six-row medication list containing a stained dose and an implausible but clearly printed quantity.
unsiloed.ai
· 2026-10-13
After TypeSafe released its Jev decision-model API, which answers yes/no or multiple-choice questions with probabilities instead of generating text, several companies including Microsoft and OpenAI released similar open-source tools. Nandakishor Mukkunnoth of ConvAI Innovations then said he had designed a nearly identical architecture over a year earlier, pointing to his March 2025 arXiv paper on SalesRLAgent, which included published model weights and a dataset.
tomshardware.com
· 2026-10-12
A technical write-up describes how constrained decoding can turn a standard language model into a fast 'decision model' that picks from a fixed set of options (A through E) in a single forward pass, rather than generating output token by token as with structured JSON output. The piece includes example code using Qwen/Qwen3-1.7B to mask the vocabulary down to allowed answer tokens and select the highest-probability one.
nishtahir.com
· 2026-10-10
TypeSafe AI, maker of the Jev model, has closed an $870 million funding round at a $7.5 billion valuation led by Andreessen Horowitz, with Sequoia and existing backer DCVC also participating. Jev launched on Sept. 15 and, per the company, is already used by a third of Fortune 500 firms. Unlike large language models, Jev uses a transformer architecture to output probabilities or 'calibrated decisions' rather than text, which TypeSafe says makes it faster and more token-efficient for automation tasks.
techcrunch.com
· 2026-10-09
Strands Labs has launched Strands Decider 2B, a small open-source 'decision model' built for fast local experimentation within its strands-labs initiative. Unlike general-purpose LLMs, it is designed to choose among predefined options or assign numerical scores rather than generate free-form text, and it also outputs a reliability score for each decision. The model is intended to integrate with the Strands Harness SDK and the recently launched Strands harness.
strandsagents.com
· 2026-10-07
The research team built a diagnosis pipeline that uses Jev, a lightweight decision-making tool, to analyze Kubernetes cluster evidence without any LLM agent involved. The pipeline programmatically collects cluster objects, events, logs, and resource data, then has Jev select likely root causes and supporting evidence to assemble a diagnosis report. Tested across 21 SREGym-Lite faults, the pipeline passed 80 of 105 diagnoses (76.2%) with a median diagnosis time of 14.6 seconds.
sregym.com
· 2026-10-07
Musubi unveiled PolicyLM-1.7B, a lightweight decision model built to moderate content in under 50 milliseconds using plain-English policies, and released it with open weights. The company says it matches the cost and speed of existing moderation classifiers but can adapt to new rules without retraining, unlike conventional systems.
techcrunch.com
· 2026-10-06
GoKawiil