Mercury 3
Mercury 3 is an API-only (closed weights) language model from Inception, released October 16, 2026. The context window is 1M tokens. API price: $0.1 input / $0.35 output per million tokens.
Today we’re introducing Mercury 3, the most capable diffusion-based LLM released to date. Across a range of benchmarks, Mercury 3 outperforms GPT-6 Luna and Claude Haiku 5.5, while running 4–8x faster at 1,000+ tokens per second. Mercury 3 demonstrates that text diffusion scales.
Check it against your own PC →
Lab
Inception
Released
October 16, 2026
Weights
API-only (closed weights)
Parameters
–
Context
1M
Price / M tokens
$0.1 / $0.35 in / out
Epoch ECI
not yet rated
Benchmark results
Epoch AI has not published results for Mercury 3 yet. New models are usually evaluated within days to weeks.
Links
More from Inception
Mercury 2.5 · Mercury Edit 2 · Mercury 2 · Mercury Coder Mini · Mercury Coder Small