{"id":"tell-1","slug":"tell-1","name":"Tell-1","orgId":"retell-ai","family":"tell-1","releaseDate":null,"availability":"ga","openWeights":false,"license":"Not stated as a weight license","paramsTotal":null,"paramsSource":null,"contextWindow":null,"knowledgeCutoff":null,"modelCardUrl":"https://www.retellai.com/changelog/tell-1","hfRepo":null,"modalities":["audio","text"],"aliases":["Tell-1","tell-1","Tell 1","retell tell-1","retell-ai tell-1","Retell Tell-1","Retell AI Tell-1"],"aliasNote":null,"origin":"tracked","tracked":{"trackedSince":"2026-11-19","announced":"2026-11-12","blurb":"Retell AI's in-house LLM for contact-center voice agents, with interleaved reasoning: it starts the reply right away and works out the next step while it speaks. Official changelog and press release are dated November 12, 2026. Official pages still describe it as in beta, running on Retell's own inference stack, and selectable as the agent LLM in the Retell dashboard; the public Create Retell LLM API lists model id tell-1. Official price is $0.035 per minute. No public weight release is stated.","api":"live","apiModelId":"tell-1","openWeights":"no","facts":[{"label":"Type","value":"Retell's in-house LLM for contact-center voice agents. Official technical blog: it writes the first sentence of a reply right away, then reasons silently (which rule applies, what is still missing, and whether to call a tool) while that sentence plays, and continues the reply grounded in that reasoning; the reasoning is never spoken (Retell changelog, Tell-1 page, and technical blog)"},{"label":"Hosted access","value":"Official changelog: select it as the LLM for your agent in the Retell dashboard. Official API docs list model id tell-1 among available LLM models on the public Create, Get, and Update Retell LLM endpoints at https://api.retellai.com, authenticated with an API key from the dashboard. Official quick start: new accounts sign up in the dashboard with $10 in free trial credits and no payment method required. Official changelog and pricing still describe Tell-1 as in beta, running on Retell's own inference stack (Retell changelog, API docs, and docs quick start)"},{"label":"Pricing","value":"Official changelog, Tell-1 page, and pricing page: $0.035 per minute. The official pricing page still labels this rate as Beta"},{"label":"Latency","value":"Official changelog and Tell-1 page: about 150 ms to first token. Official technical blog, as model-only times that exclude speech recognition and synthesis: 72 ms median time to the first full sentence (time to first meaningful chunk) over live traffic on the Retell vCX Benchmark (Retell changelog, Tell-1 page, and technical blog)"},{"label":"Weights","value":"No public weight release stated. Official changelog and Tell-1 page: it runs on Retell's own inference stack"},{"label":"License","value":"Not stated as a weight license. Access is under Retell's dashboard and API terms"}],"claims":["Lab claim: self-reported average score of 94.4% on the Retell vCX benchmark across instruction following, grounding, tool calling, and colloquial speech, with official skill scores of 88.6% instruction following, 97.2% grounding, 95.4% tool calling, and 96.6% colloquial speech (official Tell-1 page and changelog).","Lab claim: self-reported as about 45% lower cost than models with similar vCX scores, at $0.035 per minute (official technical blog).","Lab claim: self-reported 99.98% of turns pause for less than 400 ms after the first sentence, and a median 3.2 ms per token under heavy load, more than 300 tokens a second, in Retell's own inference measurements that exclude network, speech recognition, and speech synthesis (official technical blog)."],"sources":[{"kind":"lab","title":"Introducing Tell-1 (Retell changelog)","url":"https://www.retellai.com/changelog/tell-1","date":"2026-11-12"},{"kind":"lab","title":"Tell-1 (Retell)","url":"https://www.retellai.com/tell-1"},{"kind":"lab","title":"Tell-1 technical blog (Retell Labs)","url":"https://www.retellai.com/tell-1-technical-blog","date":"2026-11-13"},{"kind":"lab","title":"Create Retell LLM API","url":"https://docs.retellai.com/api-references/create-retell-llm"},{"kind":"lab","title":"Retell pricing","url":"https://www.retellai.com/pricing"},{"kind":"lab","title":"Build your first phone agent (Retell docs)","url":"https://docs.retellai.com/get-started/quick-start"},{"kind":"lab","title":"Retell launches Tell-1 (Business Wire press release)","url":"https://www.businesswire.com/news/home/20261112258243/en/Retell-Launches-Tell-1-a-Language-Model-Built-for-the-Complexity-of-Live-Contact-Center-Calls","date":"2026-11-12"}],"firstBenchmark":null},"organization":{"id":"retell-ai","name":"Retell AI","slug":"retell-ai","country":null,"url":"https://www.retellai.com","color":"#11110F","mark":"R"},"price":null,"answer":{"word":"Pending benchmarks.","scope":"Announced Nov 12, 2026; no independent results yet.","lead":"As of Nov 22, 2026, Tell-1 is pending benchmarks on isitsota: announced Nov 12, 2026, with no independent results on the boards we track yet.","confidence":"low"},"status":"PENDING_BENCHMARKS","verdicts":[]}