{"version":1,"type":"story","url":"https://digestai.news/story/typesafe-s-jev-decision-model-costs-0-042-per-million-tokens-matches-s","json":"https://digestai.news/story/typesafe-s-jev-decision-model-costs-0-042-per-million-tokens-matches-s.json","markdown":"https://digestai.news/story/typesafe-s-jev-decision-model-costs-0-042-per-million-tokens-matches-s.md","slug":"typesafe-s-jev-decision-model-costs-0-042-per-million-tokens-matches-s","headline":"TypeSafe's Jev decision model costs $0.042 per million tokens, matches Sonnet 5 accuracy","summary":"TypeSafe AI’s Jev decision model returns typed answers—choices, scores or binary probabilities—rather than generated text. It supports up to 255 options per choice and charges $0.042 per million input tokens with free output, offering a 32 K context window and response times of 70‑500 ms.\n\nIn TypeSafe’s internal workflow benchmarks across security, agent observability, invoice processing and customer service, Jev achieved 67.8 % mean accuracy at $0.0004 per case and 0.4 s latency, matching Claude Sonnet 5’s accuracy while costing far less. The top frontier model (“sol”) still led with 74.1 % accuracy but was slower and pricier. Early adopters such as Vercel AI Gateway saw nearly 13 % of paid teams use Jev within 24 hours, double the share of GPT‑5.6 and six times that of Fable 5.1.\n\nFastino Labs quickly followed with GLiDE (released Sep 30) and GLiNER2.5‑Decide (Sep 24), while open‑weight reproductions like kev, OpenJev and JevK5 allow local deployment. Decision models are positioned for high‑volume, bounded judgments—routing, classification, triage, guardrails and evaluation—while traditional LLMs remain suited for open‑ended generation.","keyPoints":["Jev costs $0.042 per million input tokens, output free.","In TypeSafe evals Jev matched Claude Sonnet 5 accuracy at 0.4 s per case.","Within 24 h of Vercel AI Gateway launch, ~13% of paid teams adopted Jev."],"whyItMatters":"Decision models like Jev give cheap, fast, calibrated answers for high‑volume tasks, letting developers replace expensive LLM calls for routing, triage, and evaluation, which lowers costs and latency across many applications.","category":{"slug":"models","name":"Generative AI & Models","url":"https://digestai.news/category/models"},"entities":{"companies":["TypeSafe AI","Fastino Labs","Vercel","OpenRouter","Arize","Langfuse"],"models":["Jev","GLiDE","GLiNER2.5-Decide","Claude Sonnet 5","GPT-6 Astra","OpenAI sol"],"people":["Simon Willison","Asif Razzaq"]},"firstPublishedAt":"2026-10-03T03:31:15Z","updatedAt":"2026-10-03T03:31:15Z","sourceCount":1,"hasPrimarySource":false,"sources":[{"outlet":"MarkTechPost","title":"Decision AI Models Explained: TypeSafe Jev vs Fastino GLiDE, GLiNER2.5-Decide and Open-Source Competitors","url":"https://marktechpost.com/2026/10/02/decision-ai-models-explained-typesafe-jev-vs-fastino-glide-gliner2-5-decide-and-open-source-competitors","publishedAt":"2026-10-03T03:31:15Z","type":"press","primary":false,"lead":true}],"sourceNotes":null,"discussions":[],"thread":null,"cite":{"text":"Digest AI, \"TypeSafe's Jev decision model costs $0.042 per million tokens, matches Sonnet 5 accuracy\", 3 October 2026, https://digestai.news/story/typesafe-s-jev-decision-model-costs-0-042-per-million-tokens-matches-s","publisher":"Digest AI","title":"TypeSafe's Jev decision model costs $0.042 per million tokens, matches Sonnet 5 accuracy","datePublished":"2026-10-03T03:31:15Z","url":"https://digestai.news/story/typesafe-s-jev-decision-model-costs-0-042-per-million-tokens-matches-s"},"generatedBy":"Written by Digest AI's editorial model from the linked sources; the sources are the record.","license":"Headlines, digests and key points are written by Digest AI and may be quoted with a link to the story page. Linked articles belong to their publishers. Terms: https://digestai.news/terms#reuse"}