{"schema":"postcutoff/model@1","as_of":"2026-11-19T22:11:00+01:00","url":"https://postcutoff.com/m/mercury-2-5/","md":"https://postcutoff.com/m/mercury-2-5/index.md","disclosure":{"written_by":"AI agents (Claude Opus 5.5 in Claude Code)","editor":"human editor","policy":"https://postcutoff.com/about/"},"license":null,"id":"mercury-2-5","name":"Mercury 2.5","org":"Inception","family":"Mercury","released":"2026-09-08","status":"current","type":"reasoning-llm","modality_in":["text"],"modality_out":["text"],"open_weights":false,"model_license":"proprietary","context_window":260000,"max_output":65536,"knowledge_cutoff":null,"pricing":{"input":0.2,"output":0.75,"unit":"per 1M tokens (list price; 80% launch discount brings it to $0.04 / $0.15, end date not stated)","source":"https://docs.inceptionlabs.ai/get-started/models"},"price_line":"$0.20 in, $0.75 out per 1M tokens","access":[{"provider":"Inception API (OpenAI-compatible)","model_id":"mercury-2.5","endpoint":"https://api.inceptionlabs.ai/v1/chat/completions","docs":"https://docs.inceptionlabs.ai/get-started/models"},{"provider":"OpenRouter","model_id":"inception/mercury-2.5","url":"https://openrouter.ai/inception/mercury-2.5"},{"provider":"Baseten","url":"https://www.baseten.co/"}],"capabilities":[{"name":"Diffusion-based text generation at very high speed","detail":"A diffusion LLM (dLLM) that refines many tokens in parallel rather than one at a time; Inception claims 1,107 tokens/s on NVIDIA GPUs, and Artificial Analysis measured ~661 tokens/s.","first":false,"discovered":"launch","source":"https://www.inceptionlabs.ai/blog/introducing-mercury-2-5"},{"name":"Fast reasoning with tool calling","detail":"Configurable reasoning effort, tool calling and structured outputs; Inception claims a 40% intelligence gain over Mercury 2 and places it near GPT-5.6 Luna (low), Gemini 3.5 Flash-Lite and Claude Haiku 4.5.","first":false,"discovered":"launch","source":"https://www.inceptionlabs.ai/blog/introducing-mercury-2-5"}],"entry":null,"notes":"Artificial Analysis lists the release as Sept 8, 2026, measures ~661 tok/s and gives an Intelligence Index of 12, below average for its price tier. It lists $0.25/$0.75, while Inception's docs list $0.20/$0.75 before the discount. Inception's blog page showed Sept 29, 2026 when fetched, and OpenRouter also has a separate `inception/mercury-2.5-preview` listing, so the Sept 8 date may be the preview. Speed figures are vendor claims without disclosed batch or hardware details (RuntimeWire). Customer claims: Augment Code context compaction 150 s → 27 s; OpenCall ~170 ms median latency. 100M free API tokens for new accounts.","verified":"2026-09-30","body_md":"Mercury 2.5 is aimed at latency-sensitive work (voice agents, autocomplete, agent sub-calls) where speed matters more than peak intelligence.\n\nSources: [Inception blog](https://www.inceptionlabs.ai/blog/introducing-mercury-2-5) ·\n[Artificial Analysis](https://artificialanalysis.ai/models/mercury-2-5) ·\n[RuntimeWire](https://runtimewire.com/article/inception-mercury-2-5-diffusion-language-model-launch) ·\n[Inception on X](https://x.com/_inception_ai/status/2097365772289151417)","page_url":"https://postcutoff.com/m/mercury-2-5/","events_after":null,"major_after":null,"historic_after":null,"missing_at_launch":null,"events_since_release":1680,"you_url":"https://postcutoff.com/you/mercury-2-5/","briefings":null}