Claude Opus 5.5 vs MiniMax M3.1 Flash
| Benchmark | Opus 5.5 | MiniMax M3.1 Flash |
|---|---|---|
| Humanity's Last Exam | 54.7 | 48 |
| AA Omniscience Index | 40.3 | 15.6 |
| CritPt (physics) | 27.7 | 22.3 |
| AA Intelligence Index 4.3.2 | 51.2 | 42.7 |
| AA Omniscience · Accuracy | 64.6 | 56.2 |
| AA Omniscience · Hallucination Rate | 68.4 | 92.8 |
| AA-LCR (long context) 1.1 | 84.3 | 79.7 |
| GDP.pdf · AA all-pass | 25.6 | 20.8 |
| MLCR-AA | 22.2 | 16.7 |
| Vals Legal Research Bench | 50.5 | 34.1 |
| Arena Text · overall | 1,508 · high | — |
| Arena Text · Chinese | 1,581 · high | — |
| Arena Text · English | 1,521.7 · high | — |
| Arena Text · French | 1,492.8 · high | — |
| Arena Text · instruction following | 1,538.8 · high | — |
| Arena Text · Japanese | 1,535.4 · high | — |
| Arena Text · non-English | 1,506.5 · high | — |
| Arena Text · Russian | 1,516.9 · high | — |
| Arena Text · Spanish | 1,509.2 · high | — |
| SimpleQA Verified v1 | 72.2 · max | — |
| ARC-AGI-1 · public evaluation | 98.3 | — |
| ARC-AGI-1 · semi-private | 97.5 | — |
| ARC-AGI-2 · public evaluation | 93.5 | — |
| ARC-AGI-2 · semi-private | 87.5 | — |
| Chess Puzzles | 67 · max | — |
| FrontierMath Erdos | 2.9 · max | — |
| FrontierMath Tier 4 (v2) | 95 · max | — |
| FrontierMath Tiers 1–3 (v2) | 91.2 · max | — |
| Mystery Game Puzzles | 71 · max | — |
| MathArena · ArXivMath · August 2026 | 86 · high | — |
| MathArena · ArXivMath · June 2026 | 83.3 · high | — |
| MathArena · ArXivMath · May 2026 | 81.7 · high | — |
| MathArena · ArXivMath · September 2026 | 82.1 · high | — |
| MathArena · BrokenArXiv · August 2026 | 77.4 · high | — |
| MathArena · BrokenArXiv · June 2026 | 100 · high | — |
| MathArena · BrokenArXiv · May 2026 | 95 · high | — |
| MathArena · BrokenArXiv · September 2026 | 76.7 · high | — |
| MathArena · ArXivLean · September 2026 | 47.2 · high | — |
| Vals MedCode | 49.8 | — |
| Vals MedScribe | 91.4 | — |
| Vals Science Index | 58.7 | — |
Specs
| Spec | Opus 5.5 | MiniMax M3.1 Flash |
|---|---|---|
| Input, per 1M | $4 | $0.15 |
| Output, per 1M | $20 | $0.60 |
| Cost per task? | $1.34 | $0.22 |
| Speed, tok/s | 82 | 121 |
| Context | 1M | 1M |
| Weights | Closed | Closed |
| Released | Sep 2026 | Oct 2026 |
| See images | ✓ | ✓ |
| Read PDFs | ✓ | — |
| Hear audio | — | — |
| Watch video | — | ✓ |
| Call tools | ✓ | ✓ |
| Force a tool call | — | ? |
| Return strict JSON | ✓ | — |
| Search the web | ✓ | ? |
| Run code | ✓ | ? |
| Use a computer | ✓ | ? |
What breaks when you switch
No PDF input
No strict JSON
Complete model response must be appended›
In multi-turn tools, preserve the complete assistant response, including reasoning_content and tool_calls, to maintain reasoning continuity.
No forced tool use›
Forced tool_choice any or tool returns HTTP 400; remove the forced choice.
Thinking is always on›
Disabled thinking or a manual budget returns HTTP 400; omit thinking or use adaptive thinking.
Remove sampling parameters›
Non-default temperature, top_p or top_k returns HTTP 400; omit these parameters and use prompting.
No assistant prefill›
Assistant prefill returns HTTP 400; use structured outputs or system instructions, with system instructions on Bedrock.
Old computer tool rejected · Claude API / Google Cloud›
The computer_20251124 tool returns HTTP 400 on the Claude API and Google Cloud; use computer_toolset_20260801 there.
Return thinking blocks unmodified›
Edited, reordered or partially dropped thinking blocks return HTTP 400; echo assistant messages complete and unmodified with tool results.