Compare / MiniMax M2.5 vs gpt-oss-20b
MiniMax M2.5 vs gpt-oss-20b
Which is cheaper, which is more capable, and which does better at each kind of work, from 16 results on 3 sources that measured both models.
MiniMax M2.5 or gpt-oss-20b? The short answer
- gpt-oss-20b is cheaper: $0.060 against $0.52 per million tokens, 8.7 times less.
- MiniMax M2.5 scores higher on the Artificial Analysis Intelligence Index: 22.8 against 10.0.
- On our business benchmarks, MiniMax M2.5 ranks higher for reasoning and knowledge and writing.
- On our business benchmarks, gpt-oss-20b ranks higher for agents and tool use.
At a glance
| MiniMax M2.5 | gpt-oss-20b | |
|---|---|---|
| Price per million tokens | $0.30 in · $1.20 out | $0.030 in · $0.15 out |
| Blended (3 in : 1 out) | $0.52 | $0.060 |
| Cheapest host, blended | $0.44 (Venice) | $0.036 (Darkbloom, FP8) |
| Intelligence Index | 22.8 | 10.0 |
| Spring Prompt overall | – | – |
| Output speed, tokens a second | – | 249 |
| Context window | 204,800 tokens | 131,072 tokens |
| Weights | Open | Open |
| Developer based in | China | the US |
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 models available today, for each kind of work, from our own benchmarks and licensed sources. The better rank is in green.
| Use case | MiniMax M2.5 | gpt-oss-20b |
|---|---|---|
| Agents and tool useMulti-step tasks with tools: support desks, coding agents, function calls | 132 of 161 | 124 of 161 |
| Reasoning and knowledgeHard questions across science, maths and general knowledge | 80 of 195 | 134 of 195 |
| WritingWhat people prefer in blind comparisons, and judged writing quality | 112 of 187 | 170 of 187 |
| Sticking to the factsSummarising without inventing things, and factual answers | 103 of 153 | not measured |
Every shared result
| Benchmark | Metric | MiniMax M2.5 | gpt-oss-20b |
|---|---|---|---|
| Arena (formerly LMArena)reported | Business, management and financeArena rating | 1,400 | 1,323 |
| Arena (formerly LMArena)reported | Creative writingArena rating | 1,357 | 1,240 |
| Arena (formerly LMArena)reported | Expert promptsArena rating | 1,425 | 1,314 |
| Arena (formerly LMArena)reported | Instruction followingArena rating | 1,381 | 1,281 |
| Arena (formerly LMArena)reported | OverallArena rating | 1,391 | 1,318 |
| Arena (formerly LMArena)reported | Writing, literature and languageArena rating | 1,369 | 1,275 |
| Artificial Analysisreported | Artificial Analysis Intelligence Indexindex score | 22.8 | 10.0 |
| Artificial Analysisreported | GPQA Diamond% of questions | 84.8% | 68.8% |
| Artificial Analysisreported | Humanity's Last Exam% of questions | 20.5% | 11.0% |
| Artificial Analysisreported | IFBench% of instructions | 71.6% | 65.1% |
| Artificial Analysisreported | Long-context reasoning (AA-LCR)% of questions | 73.3% | 34.7% |
| Artificial Analysisreported | Terminal-Bench Hard% of tasks | 34.8% | 10.6% |
| Artificial Analysisreported | Τ²-bench telecom% of tasks | 95.3% | 60.2% |
| UGI Leaderboardreported | Requested-length error% off the requested word count | 41.0% | 98.0% |
| UGI Leaderboardreported | Style adherencescore from 0 to 1 | 0.38 | 0.31 |
| UGI Leaderboardreported | Writing scorescore out of 100 | 41.7 | 24.6 |
Bold green marks the better value on that metric. Where a source reports ranges that overlap, the difference may not be meaningful; see the benchmark page for ranges.
Quick answers
Is MiniMax M2.5 better than gpt-oss-20b?
It depends on the task. Among models available today, on our benchmarks and licensed sources, MiniMax M2.5 ranks higher for reasoning and knowledge and writing; gpt-oss-20b ranks higher for agents and tool use.
Which is cheaper, MiniMax M2.5 or gpt-oss-20b?
gpt-oss-20b costs $0.060 per million tokens (three input to one output), against $0.52 for MiniMax M2.5.
Run MiniMax M2.5 and gpt-oss-20b on your own prompt
Benchmarks aren't your data. Try both side by side in the playground, with the cost of every answer.