← Back to explorer
MiniMaxReleased 2026-04Open weights

MiniMax M2.5

Open-weight coding contender hitting ~80% SWE-bench Verified — proof the open tier has closed much of the engineering gap.

Context
200K
Speed
80 t/s
Price /M
$0.30 / $1.2
Access
open weights

When to use it

Open coding agents and competitive self-hosted SWE workflows.

Watch-outs

Broader Arena/AA coverage still thinner than the biggest labs.

CodingBest valueAgents

Benchmarks

Arena Elo

Human preference ranking from blind pairwise chats. Higher is better; top frontier models cluster within ~50–80 Elo.

AA Intelligence Index44

Artificial Analysis composite across agents, coding, science, and general evaluations (v4.1 weighting).

SWE-bench Verified80.2%

Percent of real GitHub issues resolved end-to-end. Strong signal for agentic coding usefulness.

GPQA Diamond

PhD-level science questions. Separates frontier reasoning models better than saturated knowledge tests.

MMLU-Pro

Harder multi-choice knowledge/reasoning suite than classic MMLU.

Humanity's Last Exam

Frontier closed-ended academic difficulty across many domains.

Model Gauge · built by Cursor Grok 4.5 High · curated snapshot, not a live leaderboard

Scores change weekly — verify critical decisions against primary sources.