Mercury 3: Diffusion at scale
What happened
Inception Labs introduced Mercury 3, a diffusion‑based large language model that claims to be the most capable of its kind, outperforming GPT‑6 Luna and Claude Haiku 5.5 while generating 1,000+ tokens per second, 4–8× faster than rivals. The model is over 40× larger than Mercury 1 and costs $0.20 per million input tokens and $0.70 per million output tokens, with a launch discount of 50 %. Early‑access partners are using it for voice and chat support, real‑time tutoring, data engineering, and web navigation, reporting near‑real‑time responses and lower per‑token costs. Tests on 50 live Mind2Web tasks showed Mercury 3 completing 50 % of them, more than the frontier model GPT‑6.1‑Sol, with median generation times under a second.
Why it matters
For ordinary readers, faster, cheaper AI responses could make interactive tools like tutoring apps or customer‑service bots feel more immediate and affordable. The speed gains also mean less waiting time in everyday tasks that rely on AI assistance.
Reality check
An automated, sceptical read of the sourceIncremental. The article presents Mercury 3 as a performance and cost improvement over existing models rather than a fundamentally new capability.
Is this a big deal?
Get stories like this every morning
The day's AI news in plain English, in your inbox at 8:30 IST. Free.
More AI news
AI unlocks Gödel's notebooks, said to contain hundreds of proved theoremsIndustry news · 13 Nov 2026Gemini 4 Hydrogen and OrchestratorIndustry news · 12 Nov 2026Gemini 4 HydrogenIndustry news · 12 Nov 2026I'm making LLMs work on small and old devicesIndustry news · 12 Nov 2026United Airlines bans AI agents from booking flights on behalf of travelersIndustry news · 11 Nov 2026This page was collected from inceptionlabs.ai and summarised automatically by AIPulsee's pipeline. See how it runs.