Mercury 2.5 Benchmarks, Pricing & Rankings
Inception Labs · 2026-06-20 · Open · Current
Snapshot 2026-10-14 · updated weekly
Mercury 2.5 ranks #39 of 42 on the AMR leaderboard.
Mercury 2.5 benchmark scores
| GPQA Diamond | 48.2% | |
| MMLU-Pro | 47.5% | |
| SWE-Bench Verified | 24.6% | |
| AIME 2025 | 41.8% | |
| LiveCodeBench v6 | 38.2% | |
| Arena Elo | 1244 |
Mercury 2.5 price, speed & context
Input price
$0.10
per 1M tokens
Output price
$0.40
per 1M tokens
Speed
583
tok/s
Context
128K
tokens
About Mercury 2.5
Mercury 2.5 is a diffusion LLM: 583 tok/s median output with usable quality, built for drafting and classification at scale.
Models around Mercury 2.5 in the rankings
Mercury 2.5 FAQ
Where does Mercury 2.5 rank?
Mercury 2.5 ranks #39 of 42 tracked AI models on the AMR aggregate as of October 2026, with an aggregate score of 62.0. Its strongest dimension is GPQA Diamond at 48.2%.
How much does Mercury 2.5 cost?
Mercury 2.5 costs $0.10 per million input tokens and $0.40 per million output tokens on the standard API. At a typical 1:3 input/output agent workload, that blends to roughly $0.33 per million tokens.
What is the Mercury 2.5 context window?
Mercury 2.5 supports a 128K token context window and generates at a median 583 output tokens per second. That context fits roughly 85 standard documents in a single prompt.
Is Mercury 2.5 open source?
Yes — Mercury 2.5 ships open weights that you can download and self-host or serve through any compatible inference provider.
What is better than Mercury 2.5?
Clef 2 ranks one place higher at #38 with an AMR score of 67.2. Mercury 2.5 still wins 0 of the tracked dimensions head-to-head — check the comparison table to see exactly which.