---
id: "inception/mercury-2"
also: ["inception/mercury-2-20260304", "vercel/inception/mercury-2"]
card_usd_per_m: "0.25 in, 0.75 out, at inception"
cheapest_list_usd_per_m: "0.25 in, 0.75 out, at inception on openrouter"
cheapest_effective_usd_per_m: "0.1713 in, 0.7486 out, at inception on openrouter"
context: 128000
max_output: 50000
input: ["text"]
output: ["text"]
released: "2026-03-04"
reasoning: ["high", "medium", "low", "none"]
api: ["/api/v1/chat/completions", "/api/v1/completions", "/api/v1/messages", "/api/v1/responses"]
params: ["include_reasoning", "max_tokens", "reasoning", "reasoning_effort", "response_format", "stop", "structured_outputs", "temperature", "tool_choice", "tools"]
endpoints: 2
gateways: "openrouter (1), vercel (1)"
no_training: 1
trains: 0
retains_prompts: 0
coding: 31.1000
agentic: 1.6000
---
# Inception: Mercury 2

Mercury 2 is an extremely fast reasoning LLM, and the first reasoning diffusion LLM (dLLM). Instead of generating tokens sequentially, Mercury 2 produces and refines multiple tokens in parallel, achieving...