VerdictPal
Join

Models / MiniMax

MiniMax M2.7

MiniMax · Released 2026-04-15

MiniMax's April 2026 mid-tier. Vals Index 41.41 (#13). Bridge between M2.5 and the M3 open-weights release.

MiniMax mid-tier when M3 is too new for your stack or M2.5 scores aren't enough.

#18 of 21 on Vals Index · current snapshot
41 Vals Index · 11 benchmark rows · 2 sourcesPosition is benchmark-specific — not a cross-family or cross-source ranking.
Composite50
Context1M
Input / 1M$0.35
Output / 1M$1.5
Knowledge cutoff2026-02-01
Output speed59 t/s
TTFT (default API)1.17s
Statussolid
ReasoningCodingAgenticLong context
Reasoning
87
Coding
47
Agentic
55
Long context
69

65average across 4 tested families

Benchmark placements

Where MiniMax M2.7 places on each public benchmark source that publishes a row. Rank counts every model with a latest row in the same test — not a universal quality score.

BenchmarkFamilyRankScoreSourceDate
IFBenchReasoning#17/ 5776Artificial Analysis2026-07-21
Vals IndexAgentic#17/ 2041Vals AI — Vals Index2026-06-04
AA-LCRLong context#29/ 6469Artificial Analysis2026-07-21
GPQA DiamondReasoning#35/ 6987Artificial Analysis2026-07-21
SciCodeCoding#36/ 6547Artificial Analysis2026-07-21
Artificial Analysis Intelligence IndexReasoning#37/ 6738Artificial Analysis2026-07-21
Humanity's Last ExamReasoning#38/ 6528Artificial Analysis2026-07-21
AA time to first tokenPerformance#41/ 69Artificial Analysis2026-07-21
Terminal-BenchAgentic#41/ 6955Artificial Analysis2026-07-21
AA output speedPerformance#49/ 69Artificial Analysis2026-07-21
τ³-Bench BankingAgentic#57/ 649Artificial Analysis2026-07-21
Family context

The three highest-scoring carded models in each capability family. Where MiniMax M2.7 shows up, it's highlighted.

ReasoningMiniMax M2.7 · 87
  1. 1
    Claude Mythos Preview
    95
  2. 2
    GPT-5.6 Sol
    94
  3. 3
    Kimi K3
    94
CodingMiniMax M2.7 · 47
  1. 1
    GPT 5.5
    96
  2. 2
    Claude Opus 4.8
    96
  3. 3
    Claude Opus 4.6
    96
AgenticMiniMax M2.7 · 55
  1. 1
    GPT 5.5
    100
  2. 2
    Claude Opus 4.8
    100
  3. 3
    Claude Fable 5
    100
Long contextMiniMax M2.7 · 69
  1. 1
    Kimi K3
    75
  2. 2
    GPT-5.1
    75
  3. 3
    GPT-5.6 Sol
    74
Newest receipts

Benchmark rows added to the public ledger for MiniMax M2.7 in the last 120 days. Older rows live in the full table below.

11 recent
BenchmarkFamilyScoreSourceDays ago
AA-LCRLong context68.67%Artificial Analysis2
AA output speedPerformance59 t/sArtificial Analysis2
AA time to first tokenPerformance1.17sArtificial Analysis2
Artificial Analysis Intelligence IndexReasoning38.1%Artificial Analysis2
GPQA DiamondReasoning87.4%Artificial Analysis2
Humanity's Last ExamReasoning28.1%Artificial Analysis2
IFBenchReasoning75.71%Artificial Analysis2
SciCodeCoding47%Artificial Analysis2
τ³-Bench BankingAgentic8.87%Artificial Analysis2
Terminal-BenchAgentic55.43%Artificial Analysis2
Vals IndexAgentic41.41%Vals AI — Vals Index49
Family coverage

A high composite that hides a weak family is a trap. These bars surface the families where this model hasn't been publicly tested, and where it leads.

Reasoning87
Coding47
Agentic55
Long context69
Price vs. performance
0255075100$0.10$0.30$1$3$10$30input USD / 1M tokens · log scale →← intelligence

Hover or tab any dot for name, score, and input price.

AnthropicOpenAIMistralAlibabaDeepSeekGoogleMoonshotMiniMaxxAIMetaZhipuXiaomiMeituanCohere

Full benchmark ledger

Every catalog benchmark for MiniMax M2.7. Scores link to the original source; gaps mean no public row exists yet.

11 sourced rows

11 of 32 catalog benchmarks have a sourced row for MiniMax M2.7.

Reasoning

Coding

Agentic

Long context

Performance

Sources