Inception launches Mercury 3, claims edge over GPT-6 Luna

Inception released Mercury 3 on Monday, October 19. The diffusion-based language model has a 1-million-token context window, and the company says it outperforms GPT-6 Luna and Claude Haiku 5.5 on a range of benchmarks while running four to eight times faster. The announcement lists a speed of 1,068 tokens per second on NVIDIA GPUs. The release is Inception's biggest test so far of whether diffusion models can grow larger without losing their speed advantage. Every performance figure so far comes from Inception itself.businesswire.com+1

Bigger model, same speed

Standard language models write text one token at a time. Diffusion models generate many tokens in parallel and then refine them. Inception says Mercury 3 is more than 40 times larger than Mercury 1 and still runs above 1,000 tokens per second. "We're showing that the limits to scaling diffusion models are compute and data, not fundamental algorithmic barriers," said co-founder Stefano Ermon, who added that training on the next model has already started.interestingengineering.com+1

Mercury 3 is the company's fourth release since Mercury 1 launched in June 2025. The context window grows from the 260,000 tokens that Mercury 2.5 supported. In company testing on 50 live Mind2Web web-navigation tasks, Inception said Mercury 3 completed half of them, which it described as well above GPT-6.1-Sol's result.unite.ai+1

Pricing and availability

List prices are $0.20 per million input tokens and $0.70 per million output tokens. A 50% launch discount brings those to $0.10 and $0.35, and cached input tokens cost 90% less than regular input. The model is available through the Inception API, Baseten, OpenRouter and Vercel, and it works with the OpenAI API format. Mercury 2 and 2.5 are no longer open to new users, and existing customers keep access until November 15. Education startup RevisionDojo said hints and marked answers now come back in about a second.businesswire.com+2

Claims await independent checks

No independent evaluator has checked Inception's benchmark comparisons yet. As of October 20, the benchmarking firm Artificial Analysis had no Mercury 3 listing. It had previously measured Mercury 2.5 at about 706 tokens per second, which was below the speed Inception claimed for that model. OpenRouter's live traffic data showed a median throughput of 221 tokens per second on October 20. That is faster than both rival models but about a fifth of Inception's headline figure. Co-founder Volodymyr Kuleshov said on Hacker News that third-party benchmarks are still to come.startupfortune.com+2