GPT-5.5 vs MiniMax M2.7
GPT-5.5 leads MiniMax M2.7 on coding, 71.3 to 36.2 on the Vibe Coding Index, and MiniMax M2.7 costs 25.0x less per output token ($1.20 vs $30.00 per million). Updated October 22, 2026; 73 benchmarks have a result for both.
GPT-5.5 or MiniMax M2.7: which to pick
Pick GPT-5.5 for coding: it scores 71.3 on the Vibe Coding Index against 36.2 for MiniMax M2.7.
Pick MiniMax M2.7 for output tokens 25.0x cheaper ($1.20 vs $30.00 per million) and a lower cost per solved task ($0.64 against $1.65).
- Shared results
- 73
- GPT-5.5 only
- 67
- MiniMax M2.7 only
- 1
- Like-for-like categories
- 8 of 9
Only benchmarks that tested both models count toward a pick; a result only one model has is shown but not compared.
Which one for your work
Each card is decided by one number both models carry. Where only one has it, the card says so and makes no call.
Coding
Vibe Coding Index, six public coding benchmarks
GPT-5.5
GPT-5.5 scores 71.3 against 36.2.
5 and 3 benchmarks behind the scores
Intelligence
AA Intelligence Index
GPT-5.5
GPT-5.5 leads, 38.4 against 22.8.
Same board, both measured
Agentic work
AA Agentic Index
GPT-5.5
GPT-5.5 leads, 36.4 against 15.3.
Same board, both measured
Reasoning
GPQA Diamond
GPT-5.5
GPT-5.5 leads, 93.2% against 86.6%.
Same board, both measured
Math
AIME
No call
Only MiniMax M2.7 has a AIME result, 91.0%.
Directional only
Knowledge
MMLU Pro
GPT-5.5
GPT-5.5 leads, 88.1% against 80.4%.
Same board, both measured
Web design
Design Arena: websites
GPT-5.5
GPT-5.5 leads, 1256 against 1242.
Same board, both measured
Long documents
Listed context window
GPT-5.5
GPT-5.5 takes 1.05M tokens against 205K.
Listed by OpenRouter
Everyday cost
1K input + 500 output tokens
MiniMax M2.7
GPT-5.5 $0.020, MiniMax M2.7 $0.0009 per turn.
List prices
Cost per solved task
One SWE-bench Verified task, solved
MiniMax M2.7
GPT-5.5 $1.65, MiniMax M2.7 $0.64.
Published attempt cost and pass rate
At a glance
- Coding index
- GPT-5.5 71.3 from 5 benchmarks, MiniMax M2.7 36.2 from 3 benchmarks.
- Widest category
- Coding, 23.1 points apart on average over 17 shared benchmarks.
- Widest single gap
- Vibe Code Bench (Vals AI): GPT-5.5 69.8% against MiniMax M2.7 11.9%, 57.9 points apart on a 0 to 100 scale.
- Output price
- GPT-5.5 $30.00 against MiniMax M2.7 $1.20 per million output tokens, 25.0x apart.
- Context window
- GPT-5.5 takes 1.05M tokens, MiniMax M2.7 205K.
Benchmark by benchmark
First the score: boards each model leads, per category, where both have a result. Then every board, a category to a fold; bars are each result on a 0 to 100 scale, so compare the two in a row, not one row with another.
| Category | GPT-5.5 | Bars | MiniMax M2.7 | Ties |
|---|---|---|---|---|
| Coding | 17 | 0 | ||
| Intelligence | 2 | 0 | ||
| Agentic work | 6 | 0 | ||
| Reasoning | 6 | 0 | ||
| Math | 3 | 0 | ||
| Knowledge | 10 | 0 | ||
| Web design | 8 | 0 | ||
| General | 21 | 0 | ||
| All shared boards | 73 | 0 |
CodingLike-for-likeSix of these boards feed the Vibe Coding Index170
| Benchmark | GPT-5.5 | MiniMax M2.7 | Leader |
|---|---|---|---|
| LiveBench coding | 85.1% | Not measured | Only GPT-5.5 measured |
| LMArena WebDev | 1513 | 1399 | GPT-5.5 |
| AA Coding Index | 74.9 | 52.6 | GPT-5.5 |
| LiveCodeBench (Vals AI) | 85.3% | 79.9% | GPT-5.5 |
| SWE-bench Verified (Vals AI) | 82.6% | 73.8% | GPT-5.5 |
| Vibe Code Bench (Vals AI) | 69.8% | 11.9% | GPT-5.5 |
| Code migration (Vals AI) | 45.2% | 8.7% | GPT-5.5 |
| ProgramBench (Vals AI) | 0.5% | 0.0% | GPT-5.5 |
| Terminal-Bench 2.1 (Vals AI) | 76.4% | 48.7% | GPT-5.5 |
| SkillsBench (Vals AI) | 62.5% | Not measured | Only GPT-5.5 measured |
| SciCode | 56.1% | 50.1% | GPT-5.5 |
| Terminal-Bench Hard (Artificial Analysis run) | 60.6% | 39.4% | GPT-5.5 |
| Terminal-Bench | 84.7% | 45.1% | GPT-5.5 |
| FrontierCode | 43.0% | Not measured | Only GPT-5.5 measured |
| GSO-Bench | 40.2% | Not measured | Only GPT-5.5 measured |
| WeirdML | 84.9% | 37.0% | GPT-5.5 |
| MirrorCode | 10.0% | Not measured | Only GPT-5.5 measured |
| ALE-Bench | 1943.0 | 599.3 | GPT-5.5 |
| LMArena Text: coding | 1494 | 1454 | GPT-5.5 |
| LMArena Text: software and IT | 1493 | 1446 | GPT-5.5 |
| LMArena WebDev: React | 1506 | 1388 | GPT-5.5 |
| LMArena WebDev: HTML | 1543 | 1401 | GPT-5.5 |
| LMArena WebDev: from an image | 1532 | Not measured | Only GPT-5.5 measured |
| SWE Atlas: Codebase QnA | 45.4% | Not measured | Only GPT-5.5 measured |
| SWE Atlas: Refactoring | 44.8% | Not measured | Only GPT-5.5 measured |
| SWE Atlas: Test Writing | 42.6% | Not measured | Only GPT-5.5 measured |
| Neither model has a result on the other 17 boards here | |||
IntelligenceLike-for-likePublic boards, shown as published, kept out of the index20
| Benchmark | GPT-5.5 | MiniMax M2.7 | Leader |
|---|---|---|---|
| AA Intelligence Index | 38.4 | 22.8 | GPT-5.5 |
| Epoch Capabilities Index | 159.7 | 145.8 | GPT-5.5 |
Agentic workLike-for-likePublic boards, shown as published, kept out of the index60
| Benchmark | GPT-5.5 | MiniMax M2.7 | Leader |
|---|---|---|---|
| AA Agentic Index | 36.4 | 15.3 | GPT-5.5 |
| Design Arena: agents building web apps | 1117 | Not measured | Only GPT-5.5 measured |
| Design Arena: agents building full-stack apps | 1089 | Not measured | Only GPT-5.5 measured |
| Design Arena: agents building mobile apps | 1142 | Not measured | Only GPT-5.5 measured |
| Design Arena: agents building Android apps | 1163 | Not measured | Only GPT-5.5 measured |
| Design Arena: agents building Godot games | 1173 | Not measured | Only GPT-5.5 measured |
| Design Arena: agents building games | 1178 | Not measured | Only GPT-5.5 measured |
| Design Arena: HTML slides | 1006 | Not measured | Only GPT-5.5 measured |
| Design Arena: PowerPoint slides | 1083 | Not measured | Only GPT-5.5 measured |
| LiveBench research | 61.6% | Not measured | Only GPT-5.5 measured |
| LiveBench browser | 48.0% | Not measured | Only GPT-5.5 measured |
| LiveBench long agentic | 44.6% | Not measured | Only GPT-5.5 measured |
| SREBench (Vals AI) | 3.8% | Not measured | Only GPT-5.5 measured |
| Finance Agent (Vals AI) | 51.8% | 27.9% | GPT-5.5 |
| Tax Agent Bench (Vals AI) | 21.8% | 6.9% | GPT-5.5 |
| Legal Research Bench (Vals AI) | 40.4% | 10.6% | GPT-5.5 |
| Excel Modeling Benchmark (Vals AI) | 64.5% | 28.4% | GPT-5.5 |
| Time Horizon Index (Vals AI) | 8.3% | Not measured | Only GPT-5.5 measured |
| τ²-Bench Telecom | 93.9% | 84.8% | GPT-5.5 |
| APEX-Agents | 55.1% | Not measured | Only GPT-5.5 measured |
| Remote Labor Index | 6.3% | Not measured | Only GPT-5.5 measured |
| OSWorld 2.0 | 13.0% | Not measured | Only GPT-5.5 measured |
| DeepResearch Bench | 54.0% | Not measured | Only GPT-5.5 measured |
| PostTrainBench | 27.2% | Not measured | Only GPT-5.5 measured |
| Vending-Bench 2 | $7,524 | Not measured | Only GPT-5.5 measured |
| MCP Atlas | 75.3% | Not measured | Only GPT-5.5 measured |
| Neither model has a result on the other 16 boards here | |||
ReasoningLike-for-likePublic boards, shown as published, kept out of the index60
| Benchmark | GPT-5.5 | MiniMax M2.7 | Leader |
|---|---|---|---|
| GPQA Diamond | 93.2% | 86.6% | GPT-5.5 |
| LiveBench reasoning | 89.7% | Not measured | Only GPT-5.5 measured |
| MMMU Pro (Vals AI) | 88.3% | Not measured | Only GPT-5.5 measured |
| GPQA Diamond (Artificial Analysis run) | 93.5% | 87.4% | GPT-5.5 |
| Humanity's Last Exam (Artificial Analysis run) | 45.8% | 29.6% | GPT-5.5 |
| GPQA Diamond (Epoch AI run) | 90.7% | Not measured | Only GPT-5.5 measured |
| ARC-AGI | 95.0% | Not measured | Only GPT-5.5 measured |
| ARC-AGI-2 | 85.0% | Not measured | Only GPT-5.5 measured |
| SimpleBench | 69.0% | Not measured | Only GPT-5.5 measured |
| Chess puzzles | 63.0% | Not measured | Only GPT-5.5 measured |
| Mystery game puzzles | 56.0% | Not measured | Only GPT-5.5 measured |
| EBR-bench | 34.3% | Not measured | Only GPT-5.5 measured |
| LMArena Text: hard prompts | 1489 | 1422 | GPT-5.5 |
| LMArena Text: expert | 1507 | 1443 | GPT-5.5 |
| LMArena Text: hard prompts in English | 1483 | 1441 | GPT-5.5 |
| LMArena Vision: diagrams | 1325 | Not measured | Only GPT-5.5 measured |
| LMArena Vision: homework | 1332 | Not measured | Only GPT-5.5 measured |
| Kagi LLM Benchmark | 88.8% | Not measured | Only GPT-5.5 measured |
| Neither model has a result on the other 5 boards here | |||
MathLike-for-likePublic boards, shown as published, kept out of the index30
| Benchmark | GPT-5.5 | MiniMax M2.7 | Leader |
|---|---|---|---|
| AIME | Not measured | 91.0% | Only MiniMax M2.7 measured |
| OTIS Mock AIME | 84.4% | Not measured | Only GPT-5.5 measured |
| FrontierMath, tiers 1 to 3 | 85.3% | Not measured | Only GPT-5.5 measured |
| FrontierMath, tier 4 | 72.5% | Not measured | Only GPT-5.5 measured |
| ProofBench | 50.0% | 3.0% | GPT-5.5 |
| LMArena Text: math | 1487 | 1421 | GPT-5.5 |
| LMArena Text: mathematical work | 1496 | 1437 | GPT-5.5 |
| MathArena | 61.3% | Not measured | Only GPT-5.5 measured |
| Neither model has a result on the other 6 boards here | |||
KnowledgeLike-for-likePublic boards, shown as published, kept out of the index100
| Benchmark | GPT-5.5 | MiniMax M2.7 | Leader |
|---|---|---|---|
| MMLU Pro | 88.1% | 80.4% | GPT-5.5 |
| LegalBench (Vals AI) | 86.5% | 84.0% | GPT-5.5 |
| CaseLaw (Vals AI) | 66.2% | 60.9% | GPT-5.5 |
| CorpFin (Vals AI) | 68.4% | 61.2% | GPT-5.5 |
| TaxEval (Vals AI) | 75.0% | 66.6% | GPT-5.5 |
| MortgageTax (Vals AI) | 68.8% | Not measured | Only GPT-5.5 measured |
| MedCode (Vals AI) | 49.1% | 34.4% | GPT-5.5 |
| SimpleQA Verified | 63.0% | Not measured | Only GPT-5.5 measured |
| LMArena Text: science | 1485 | 1416 | GPT-5.5 |
| LMArena Text: medicine and healthcare | 1465 | 1409 | GPT-5.5 |
| LMArena Text: legal and government | 1491 | 1403 | GPT-5.5 |
| LMArena Text: business and finance | 1478 | 1407 | GPT-5.5 |
| FACTS Benchmark Suite | 66.7% | Not measured | Only GPT-5.5 measured |
| Neither model has a result on the other 7 boards here | |||
Web designLike-for-likePublic boards, shown as published, kept out of the index80
| Benchmark | GPT-5.5 | MiniMax M2.7 | Leader |
|---|---|---|---|
| Design Arena: websites | 1256 | 1242 | GPT-5.5 |
| Design Arena: UI components | 1257 | 1207 | GPT-5.5 |
| Design Arena: game dev | 1293 | 1209 | GPT-5.5 |
| Design Arena: data viz | 1264 | 1233 | GPT-5.5 |
| Design Arena: 3D | 1197 | 1194 | GPT-5.5 |
| Design Arena: all code categories | 1259 | 1233 | GPT-5.5 |
| Design Arena: SVG | 1233 | 1137 | GPT-5.5 |
| Design Arena: ASCII art | 1255 | 1136 | GPT-5.5 |
| LiveBench design | 70.3% | Not measured | Only GPT-5.5 measured |
GeneralLike-for-likePublic boards, shown as published, kept out of the index210
| Benchmark | GPT-5.5 | MiniMax M2.7 | Leader |
|---|---|---|---|
| LiveBench data analysis | 81.6% | Not measured | Only GPT-5.5 measured |
| Public Benefits Bench (Vals AI) | 60.9% | Not measured | Only GPT-5.5 measured |
| SAGE (Vals AI) | 51.5% | Not measured | Only GPT-5.5 measured |
| MedScribe (Vals AI) | 86.9% | 79.9% | GPT-5.5 |
| Vals Multimodal Index | 68.1% | Not measured | Only GPT-5.5 measured |
| IFBench | 75.9% | 75.7% | GPT-5.5 |
| Long context reasoning (AA-LCR) | 84.3% | 78.3% | GPT-5.5 |
| DTBench | 96.0% | Not measured | Only GPT-5.5 measured |
| LMCA | 54.3% | Not measured | Only GPT-5.5 measured |
| Surface Evolver Bench | 88.1% | Not measured | Only GPT-5.5 measured |
| Furniture assembly | 44.2% | Not measured | Only GPT-5.5 measured |
| ExploitBench | 47.4% | 13.3% | GPT-5.5 |
| CL-bench Life | 22.2% | Not measured | Only GPT-5.5 measured |
| LMArena Text | 1471 | 1405 | GPT-5.5 |
| LMArena Text: instruction following | 1478 | 1405 | GPT-5.5 |
| LMArena Text: creative writing | 1455 | 1354 | GPT-5.5 |
| LMArena Vision | 1297 | Not measured | Only GPT-5.5 measured |
| LMArena Text: English | 1470 | 1426 | GPT-5.5 |
| LMArena Text: non-English | 1467 | 1382 | GPT-5.5 |
| LMArena Text: Chinese | 1534 | 1442 | GPT-5.5 |
| LMArena Text: German | 1480 | 1398 | GPT-5.5 |
| LMArena Text: Japanese | 1499 | 1260 | GPT-5.5 |
| LMArena Text: Korean | 1460 | 1314 | GPT-5.5 |
| LMArena Text: French | 1485 | 1424 | GPT-5.5 |
| LMArena Text: Spanish | 1468 | 1402 | GPT-5.5 |
| LMArena Text: Russian | 1473 | 1383 | GPT-5.5 |
| LMArena Text: Polish | 1465 | 1387 | GPT-5.5 |
| LMArena Text: multi-turn | 1476 | 1412 | GPT-5.5 |
| LMArena Text: longer queries | 1484 | 1418 | GPT-5.5 |
| LMArena Text: writing and language | 1466 | 1376 | GPT-5.5 |
| LMArena Text: entertainment and media | 1449 | 1360 | GPT-5.5 |
| LMArena Vision: English | 1292 | Not measured | Only GPT-5.5 measured |
| LMArena Vision: Chinese | 1360 | Not measured | Only GPT-5.5 measured |
| LMArena Vision: reading text | 1306 | Not measured | Only GPT-5.5 measured |
| LMArena Vision: humor | 1293 | Not measured | Only GPT-5.5 measured |
| LMArena Vision: recognizing things | 1231 | Not measured | Only GPT-5.5 measured |
| LMArena Vision: creative writing | 1301 | Not measured | Only GPT-5.5 measured |
| LMArena Document | 1486 | Not measured | Only GPT-5.5 measured |
| EQ-Bench 4 | 1315 | Not measured | Only GPT-5.5 measured |
| Neither model has a result on the other 9 boards here | |||
Image generationNot comparablePublic boards, shown as published, kept out of the index0 boards, none shared
| Benchmark | GPT-5.5 | MiniMax M2.7 | Leader |
|---|---|---|---|
| Neither model has a result on 5 boards here | |||
What each workload costs
Three fixed token mixes at each model's listed price, so per-token rates become a bill.
Chat turn
1K input + 500 output tokens
MiniMax M2.7 costs less
Repository review
50K input + 3K output tokens
MiniMax M2.7 costs less
Agent loop
200K input + 10K output tokens
MiniMax M2.7 costs less
Listed OpenRouter rates, without cache or batch discounts. A workload that does not fit a model's context window would need splitting.
Specification differences
The leading value in a row is marked; a row with a missing value names no leader. Our two rows are marked in red.
| Specification | OpenAI GPT-5.5 | MiniMax M2.7 |
|---|---|---|
| Maker | OpenAI | MiniMax |
| Weights | No public weights listed | Open weights |
| Released | April 24, 2026 | March 18, 2026 |
| Context window | 1.05M | 205K |
| Input $/Mtok | $5.00 | $0.30 |
| Output $/Mtok | $30.00 | $1.20 |
| Coding index | 71.3 | 36.2 |
| Independent measurements behind it | 5 | 3 |
| Cost per solved task | $1.65 | $0.64 |
| Source | OpenRouter listing | OpenRouter listing |
Questions
Is GPT-5.5 better than MiniMax M2.7 for coding?
Pick GPT-5.5 for coding: it scores 71.3 on the Vibe Coding Index against 36.2 for MiniMax M2.7. Pick MiniMax M2.7 for output tokens 25.0x cheaper ($1.20 vs $30.00 per million) and a lower cost per solved task ($0.64 against $1.65).
Which is cheaper, GPT-5.5 or MiniMax M2.7?
GPT-5.5 costs $5.00 per million input tokens and $30.00 per million output tokens; MiniMax M2.7 costs $0.30 and $1.20. Per solved SWE-bench Verified task, GPT-5.5 costs $1.65 and MiniMax M2.7 $0.64.
Which has the bigger context window, GPT-5.5 or MiniMax M2.7?
GPT-5.5, with 1.05M tokens against 205K.
How many benchmarks is the GPT-5.5 vs MiniMax M2.7 comparison based on?
73 benchmarks have a result for both models as of October 22, 2026. GPT-5.5's coding index rests on 5 independent benchmarks and MiniMax M2.7's on 3. Where only one model has a result, the page says so instead of naming a winner.