MiniMax M2.7 minimax-m2.7

minimax · minimax reasoning agentic efficient

unranked — no quality signals yet

params
229B (A10B)
arch
moe
context
200k
license
other open weights
released
Apr 2026
reasoning
yes
downloads/30d
1.2M

# why ranked

Overall: provisional — score 38.1 ● low — a single signal; treat with caution

provisional: too few independent signals for a numbered position — the score is shown, but this model sorts after every ranked model.

signalweightinput (0–100)
aa_intelligence 0.50 38.1
arena_elo 0.30
bench_composite 0.20

missing signals are dropped and the remaining weights renormalized — never imputed.

Coding: provisional — score 69.0 ● low — a single signal; treat with caution

provisional: too few independent signals for a numbered position — the score is shown, but this model sorts after every ranked model.

signalweightinput (0–100)
aa_coding 0.25 69.0
aider_polyglot 0.25
arena_elo_code 0.15
swe_bench_verified 0.35

missing signals are dropped and the remaining weights renormalized — never imputed.

full methodology

# trend

collecting history — trends appear after a few daily runs.

# benchmarks

benchmarkscoresourcedate
AA Coding Index 52.6 Artificial Analysis
AA Intelligence Index 38.1 Artificial Analysis
Gdpval-aa 1178.0 / 3000 LLM Stats

# where to run

providerquantctx$/M in$/M out$/M blendtpsuptime
Maraunknown 192k $0.18$0.72$0.32 91.3%
DeepInfrafp8 192k $0.25$1.00$0.44 100.0%
Novitafp8 200k $0.27$1.08$0.47 98.4%
Morphunknown 192k $0.28$1.20$0.51
AtlasCloudfp8 192k $0.30$1.20$0.52 94.7%
Fireworksunknown 192k $0.30$1.20$0.52 100.0%
GMICloudfp8 192k $0.30$1.20$0.52 99.1%
Minimaxfp8 200k $0.30$1.20$0.52 99.2%
Togetherfp4 192k $0.30$1.20$0.52 99.9%
SambaNovaunknown 192k $0.30$1.50$0.60 99.3%
DeepInfrafp8 192k $0.40$2.00$0.80
Groqunknown 192k $0.60$1.80$0.90 99.5%
Minimaxfp8 200k $0.60$2.40$1.05
SambaNovaunknown 192k $0.60$2.40$1.05 99.1%

blended = (3·input + output) / 4 per 1M tokens · cheapest first

source aliases
llmstats
minimax-m2.7
openrouter
minimax/minimax-m2.7