Compare / MiniMax-M2.5 vs MiniMax-M2.7

MiniMax-M2.5 vs MiniMax-M2.7

Which is cheaper, which is more capable, and which does better at each kind of work, from 25 results on 5 sources that measured both models.

MiniMaxMiniMax-M2.5 Released 12 Feb 2026 · $0.29 in · $1.20 out
MiniMaxMiniMax-M2.7 Released 18 Mar 2026 · $0.30 in · $1.20 out
MiniMax-M2.5 wins 52 too close to callMiniMax-M2.7 wins 18

Of the 25 results both models have, the better value on each. Results whose reported ranges overlap are too close to call.

Results as of 21 November 2026 (catalogue release 2026-11-21-af3a2faf6a5c).

MiniMax-M2.5 or MiniMax-M2.7? The short answer

MiniMax-M2.7 is the newer version; they score about the same at about the same price.

  • They cost about the same: $0.52 and $0.52 per million tokens (three input to one output).
  • They score about the same on the Artificial Analysis Intelligence Index (22.8 and 22.8).
  • MiniMax-M2.7 is newer: released 18 Mar 2026, against 12 Feb 2026 for MiniMax-M2.5.
  • On third-party benchmarks, on the results both models have: MiniMax-M2.5 does better for agents and tool use; MiniMax-M2.7 for reasoning and knowledge.

Head to head

22.8
Intelligence Index
22.8
$0.52
Price per million tokens, blendedlower is better
$0.52
200,000
Context window, tokens
196,608

Every fact side by side

MiniMax-M2.5MiniMax-M2.7
Price per million tokens$0.29 in · $1.20 out$0.30 in · $1.20 out
Blended (3 in : 1 out)$0.52$0.52
Cheapest host, blended$0.44 (Venice)$0.37 (GMICloud, FP8)
Cached input, per million–$0.060 cached input
Intelligence Index22.822.8
Spring Prompt overall––
Output speed, tokens a second––
Context window200,000 tokens196,608 tokens
Released12 Feb 202618 Mar 2026 (newer)
WeightsOpenOpen
Developer based inChinaChina

Prices are the developer's list price where we have read it, otherwise the typical host on OpenRouter; see every model's price. Intelligence and speed from Artificial Analysis.

Which is better for what

Each model's rank among the models on sale today, for each kind of work, from our own benchmarks and licensed sources. Further right is better; the better rank is in green.

Agents and tool useMulti-step tasks with tools: support desks, coding agents, function calls 135 of 159 109 of 159
Reasoning and knowledgeHard questions across science, maths and general knowledge 91 of 187 80 of 187
WritingWhat people prefer in blind comparisons, and judged writing quality 107 of 169 104 of 169
Sticking to the factsSummarising without inventing things, and factual answers 101 of 139 117 of 139

Every shared result

Each model's best result on every measure both have, grouped by benchmark, with the setting that produced it. Under each measure, every model that benchmark has measured is a grey tick, best to the right, so you can see where the two sit in the field. The headline measures of each benchmark come first; the rest are under "Show all". Green marks the better value; where a source reports ranges that overlap, neither is marked.

MeasureMiniMax-M2.5MiniMax-M2.7
Arena (formerly LMArena) reported by Arena (formerly LMArena) · 3 results
Overall · TextArena rating, higher is better 150 · 118 of 387 measured 1,391 1,415
Overall · Text factualityArena rating, higher is better 141 · 110 of 163 measured 1,386 1,422
Business, management and finance · TextArena rating, higher is better 142 · 101 of 380 measured 1,400 1,424
Artificial Analysis reported by Artificial Analysis · 3 results
Artificial Analysis Intelligence Indexindex score, higher is better 92 · 92 of 249 measured 22.8 22.8
GPQA Diamond% of questions, higher is better 81 · 64 of 232 measured 84.8% 87.4%
Humanity's Last Exam% of questions, higher is better 116 · 83 of 250 measured 20.5% 29.6%
OpenHands Index reported by OpenHands Index · 3 results
Average% resolved, higher is better 22 · 25 of 34 measured 45.2%openhands 43.4%openhands
Front end (SWE-Bench Multimodal)% resolved, higher is better 27 · 24 of 34 measured 25.0%openhands 27.9%openhands
Greenfield (Commit0)% resolved, higher is better 25 · 21 of 34 measured 12.5%openhands 18.8%openhands
UGI Leaderboard reported by UGI Leaderboard · 3 results
Writing score · UGIscore out of 100, higher is better 103 · 120 of 230 measured 41.7 37.6
Requested-length error · UGI% off the requested word count, lower is better 192 · 182 of 230 measured 41.0% 36.0%
Style adherence · UGIscore from 0 to 1, higher is better 35 · 26 of 230 measured 0.38 0.38
Vectara Hallucination Leaderboard reported by Vectara Hallucination Leaderboard · 2 results
Hallucination rate · Vectara% of summaries, lower is better 46 · 84 of 104 measured 9.1% 12.9%
Answer rate · Vectara% of documents, higher is better 76 · 55 of 104 measured 98.2% 99.4%
Show all 25 results (11 more)
MeasureMiniMax-M2.5MiniMax-M2.7
Arena (formerly LMArena) reported by Arena (formerly LMArena) · 4 results
Creative writing · TextArena rating, higher is better 147 · 139 of 385 measured 1,357 1,364
Expert prompts · TextArena rating, higher is better 130 · 93 of 338 measured 1,425 1,453
Instruction following · TextArena rating, higher is better 147 · 114 of 387 measured 1,381 1,407
Writing, literature and language · TextArena rating, higher is better 146 · 127 of 386 measured 1,369 1,387
Artificial Analysis reported by Artificial Analysis · 4 results
IFBench% of instructions, higher is better 43 · 22 of 195 measured 71.6% 75.7%
Long-context reasoning (AA-LCR)% of questions, higher is better 98 · 59 of 241 measured 73.3% 78.3%
Terminal-Bench Hard% of tasks, higher is better 49 · 36 of 190 measured 34.8% 39.4%
τ²-bench telecom% of tasks, higher is better 20 · 60 of 193 measured 95.3% 84.8%
OpenHands Index reported by OpenHands Index · 3 results
Information gathering (GAIA)% resolved, higher is better 24 · 31 of 34 measured 47.9%openhands 25.5%openhands
Issue resolution (SWE-Bench)% resolved, higher is better 23 · 10 of 34 measured 72.6%openhands 75.6%openhands
Testing (SWT-Bench)% resolved, higher is better 18 · 15 of 34 measured 68.1%openhands 69.1%openhands

Quick answers

Is MiniMax-M2.5 better than MiniMax-M2.7?

MiniMax-M2.7 is the newer version; they score about the same at about the same price.

Which is cheaper, MiniMax-M2.5 or MiniMax-M2.7?

They cost about the same: $0.52 and $0.52 per million tokens (three input to one output).

Which is better for agents and tool use, MiniMax-M2.5 or MiniMax-M2.7?

MiniMax-M2.7 ranks 109 of 159 models on sale for multi-step tasks with tools: support desks, coding agents, function calls, against 135 for MiniMax-M2.5.

Which is better for reasoning and knowledge, MiniMax-M2.5 or MiniMax-M2.7?

MiniMax-M2.7 ranks 80 of 187 models on sale for hard questions across science, maths and general knowledge, against 91 for MiniMax-M2.5.

Which is better for writing, MiniMax-M2.5 or MiniMax-M2.7?

MiniMax-M2.7 ranks 104 of 169 models on sale for what people prefer in blind comparisons, and judged writing quality, against 107 for MiniMax-M2.5.

Which is better for sticking to the facts, MiniMax-M2.5 or MiniMax-M2.7?

MiniMax-M2.5 ranks 101 of 139 models on sale for summarising without inventing things, and factual answers, against 117 for MiniMax-M2.7.

Which has the bigger context window, MiniMax-M2.5 or MiniMax-M2.7?

MiniMax-M2.5: 200,000 tokens, against 196,608 for MiniMax-M2.7.

Run MiniMax-M2.5 and MiniMax-M2.7 on your own prompt.

Benchmarks aren't your task. Test both side by side on your prompt, with every answer checked against your rules and the cost of each one.