MiniMax M3 — API Pricing & Benchmarks

MiniMax M3 is MiniMax's flagship reasoning model, released in 2026-06. The API costs $0.30 per 1M input tokens and $1.20 per 1M output tokens ($0.52 blended at a 3:1 ratio), with cached input at $0.06. It supports a 1M-token context window. On the independently measured Artificial Analysis Intelligence Index it scores 45.4, ranking #34 of 68 models tracked on this site, with a median output speed of 85 tokens/sec.

Explore provider routes and additional benchmark sources for MiniMax M3

ProviderMiniMax
Input price / 1M tokens$0.30
Output price / 1M tokens$1.20
Cached input / 1M tokens$0.06
Blended price (3:1)$0.52
Context window1M
Max output tokens131K
Knowledge cutoff
Licenseopen weights
Input modalitiesT+I+V
Output speed (tok/s)85
Released2026-06
Artificial Analysis Intelligence Index45.4
LMArena Elo1435
GPQA Diamond92.9%
SWE-bench Verified80.5%
Humanity's Last Exam39%
MMMU78.1%

Where MiniMax M3 fits

On a blended 3:1 basis MiniMax M3 costs $0.52 per 1M tokens, cheaper than 79% of the 80 priced models in this index. As MiniMax's flagship it is aimed at the hardest workloads — complex reasoning, agentic coding and multi-step tool use — where output quality dominates cost. The weights are open, so it can be self-hosted for data-residency or cost reasons instead of consumed via API. Cached input is priced at $0.06 (80% below standard input), which rewards prompt structures with long stable prefixes — see our caching guide.

Price history

PeriodInput $/1MOutput $/1MCached $/1M
2026-07-18 → current$0.30$1.20$0.06

No list-price changes recorded since tracking began. Full dataset: price-history.json (CC BY 4.0).

Alternatives at a similar price

About MiniMax

MiniMax ships open-weights M-series models that undercut most of the market: its flagship M3 costs less than many providers' small models while posting mid-tier benchmark scores.

More MiniMax models

← Full comparison table (60+ models, live filters)