OpenRouter posted that Mercury 2.5 Preview from Inception AI is now live exclusively on its platform. OpenRouter claims it reaches 1,107 tokens per second through parallel token generation, with tunable reasoning, parallel tool calls, and schema-aligned JSON for latency-sensitive workloads. Inception announced 80 percent off pricing through September 7. Co-founders and academics from Inception amplified the launch on X. Based on the visible replies from a limited analyzed reply set, replies on X largely praised the speed for agentic tasks while a smaller share questioned quality relative to alternatives like Gemma.