# The Decoder — Inception's Mercury 3 diffusion model claims to beat GPT-6 Luna and Claude Haiku 5.5 at a fraction of the wait

- Company: The Decoder (the-decoder.com)
- Announced: 2026-10-21T14:45:45+00:00
- Category: new-model
- Coverage: not counted
- Announcement: yes
- Group: models
- Source: https://the-decoder.com/inceptions-mercury-3-diffusion-model-claims-to-beat-gpt-6-luna-and-claude-haiku-5-5-at-a-fraction-of-the-wait/
- Record: https://forck.live/items/25984-inception-s-mercury-3-diffusion-model-claims-to-beat-gpt-6-luna-and-claude
- Subject: Reporting
- Models affected: Mercury 3, Mercury 1, Mercury 1 mini, Mercury 2, GPT-6 Luna, Claude Haiku 5.5, GPT-6.1-Sol, DiffusionGemma, Gemma 4
- Context window: 1 million tokens
- Pricing: Regular pricing is $0.20 per million input tokens and $0.70 per million output tokens. Inception is cutting prices in half at launch.

AI startup Inception has released Mercury 3, a diffusion-based language model that refines a rough draft of the whole answer instead of writing word by word, which the company says makes it four to eight times faster than standard models. Inception calls Mercury 3 the most capable diffusion language model released so far, and says it generates 1,068 tokens per second on Nvidia GPUs, against about 240 for Claude Haiku 5.5 and 130 for GPT-6 Luna in the company's comparison. Inception's own benchmark table is more mixed: Mercury 3 wins eight of 12 tests at medium settings, leading IFBench with 76.84 points and SWE-Bench Atlas with 51.08, while Claude Haiku 5.5 takes the longer coding test DeepSWE and GPT-6 Luna edges ahead in two Artificial Analysis tests. No independent tests exist yet. The model handles up to 1 million tokens at once and uses an OpenAI-compatible API, and is available on Inception's own platform as well as OpenRouter, Baseten, and Vercel. Regular pricing is $0.20 per million input tokens and $0.70 per million output tokens, with Inception halving prices at launch and not saying when the discount ends. Early-access partners report speed gains: Octen AI says a smart-glasses search that took about 80 milliseconds for the search itself saw a median response of 0.9 seconds with Mercury 3, and TinyFish AI says it solved 25 of 50 real-website navigation tasks, beating GPT-6.1-Sol.

## Evidence

Verbatim from https://the-decoder.com/inceptions-mercury-3-diffusion-model-claims-to-beat-gpt-6-luna-and-claude-haiku-5-5-at-a-fraction-of-the-wait/:

> According to Inception, Mercury 3 generates 1,068 tokens per second on Nvidia GPUs. In the company's comparison, Claude Haiku 5.5 manages about 240 and GPT-6 Luna about 130. No independent tests exist yet.

---

Record: https://forck.live/items/25984-inception-s-mercury-3-diffusion-model-claims-to-beat-gpt-6-luna-and-claude
Catalogue: https://forck.live/llms.txt
Current issue: https://forck.live/feed.md