MiniMax M3 vs Qwen3.8 Max (0902)
Qwen3.8 Max (0902) leads MiniMax M3 on coding, 62.4 to 39.2 on the Vibe Coding Index, and MiniMax M3 costs 5.0x less per output token ($1.20 vs $6.00 per million). Updated October 22, 2026; 102 benchmarks have a result for both.
MiniMax M3 or Qwen3.8 Max (0902): which to pick
Pick Qwen3.8 Max (0902) for coding: it scores 62.4 on the Vibe Coding Index against 39.2 for MiniMax M3, though on 5 benchmarks against 6.
Pick MiniMax M3 for output tokens 5.0x cheaper ($1.20 vs $6.00 per million) and a lower cost per solved task ($0.56 against $1.32).
- Shared results
- 102
- MiniMax M3 only
- 15
- Qwen3.8 Max (0902) only
- 12
- 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
Qwen3.8 Max (0902)
Qwen3.8 Max (0902) scores 62.4 against 39.2.
6 and 5 benchmarks behind the scores
Intelligence
AA Intelligence Index
Qwen3.8 Max (0902)
Qwen3.8 Max (0902) leads, 45.4 against 29.2.
Same board, both measured
Agentic work
AA Agentic Index
Qwen3.8 Max (0902)
Qwen3.8 Max (0902) leads, 56.0 against 29.5.
Same board, both measured
Reasoning
GPQA Diamond
Qwen3.8 Max (0902)
Qwen3.8 Max (0902) leads, 93.7% against 92.7%.
Same board, both measured
Knowledge
MMLU Pro
Qwen3.8 Max (0902)
Qwen3.8 Max (0902) leads, 88.6% against 84.2%.
Same board, both measured
Web design
Design Arena: websites
Qwen3.8 Max (0902)
Qwen3.8 Max (0902) leads, 1275 against 1257.
Same board, both measured
Long documents
Listed context window
MiniMax M3
MiniMax M3 takes 1.05M tokens against 1M.
Listed by OpenRouter
Everyday cost
1K input + 500 output tokens
MiniMax M3
MiniMax M3 $0.0009, Qwen3.8 Max (0902) $0.0050 per turn.
List prices
Cost per solved task
One SWE-bench Verified task, solved
MiniMax M3
MiniMax M3 $0.56, Qwen3.8 Max (0902) $1.32.
Published attempt cost and pass rate
At a glance
- Coding index
- MiniMax M3 39.2 from 6 benchmarks, Qwen3.8 Max (0902) 62.4 from 5 benchmarks.
- Widest category
- Math, 22.0 points apart on average over 4 shared benchmarks.
- Widest single gap
- Terminal-Bench 4 (Vals AI): MiniMax M3 1.0% against Qwen3.8 Max (0902) 34.3%, 44.4 points apart on a 0 to 100 scale.
- Output price
- MiniMax M3 $1.20 against Qwen3.8 Max (0902) $6.00 per million output tokens, 5.0x apart.
- Context window
- MiniMax M3 takes 1.05M tokens, Qwen3.8 Max (0902) 1M.
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 | MiniMax M3 | Bars | Qwen3.8 Max (0902) | Ties |
|---|---|---|---|---|
| Coding | 3 | 16 | ||
| Intelligence | 0 | 2 | ||
| Agentic work | 1 | 14 | ||
| Reasoning | 1 | 12 | ||
| Math | 0 | 4 | ||
| Knowledge | 4 | 6 | ||
| Web design | 0 | 8 | ||
| General | 2 | 29 | ||
| All shared boards | 11 | 91 |
CodingLike-for-likeSix of these boards feed the Vibe Coding Index316
| Benchmark | MiniMax M3 | Qwen3.8 Max (0902) | Leader |
|---|---|---|---|
| LiveBench coding | 70.9% | 60.3% | MiniMax M3 |
| LMArena WebDev | 1482 | 1674 | Qwen3.8 Max (0902) |
| AA Coding Index | 58.6 | 76.2 | Qwen3.8 Max (0902) |
| LiveCodeBench (Vals AI) | 82.2% | 87.8% | Qwen3.8 Max (0902) |
| SWE-bench Verified (Vals AI) | 75.0% | 85.6% | Qwen3.8 Max (0902) |
| Terminal-Bench 4 (Vals AI) | 1.0% | 34.3% | Qwen3.8 Max (0902) |
| Vibe Code Bench (Vals AI) | 47.6% | 64.7% | Qwen3.8 Max (0902) |
| Code migration (Vals AI) | 19.9% | 24.0% | Qwen3.8 Max (0902) |
| ProgramBench (Vals AI) | Not measured | 0.0% | Only Qwen3.8 Max (0902) measured |
| IOI (Vals AI) | Not measured | 68.9% | Only Qwen3.8 Max (0902) measured |
| Terminal-Bench 2.1 (Vals AI) | 53.6% | 67.4% | Qwen3.8 Max (0902) |
| SWE-bench Multimodal (Vals AI) | 33.3% | 42.5% | Qwen3.8 Max (0902) |
| Vibe Code Bench 1-100 (Vals AI) | 9.2% | 12.8% | Qwen3.8 Max (0902) |
| SkillsBench (Vals AI) | 51.5% | 42.0% | MiniMax M3 |
| CyberBench (Vals AI) | 65.0% | 28.6% | MiniMax M3 |
| SciCode | 47.1% | 53.2% | Qwen3.8 Max (0902) |
| Terminal-Bench Hard (Artificial Analysis run) | 42.4% | Not measured | Only MiniMax M3 measured |
| FrontierCode | 14.7% | Not measured | Only MiniMax M3 measured |
| FrontierSWE | Not measured | 17.8% | Only Qwen3.8 Max (0902) measured |
| ALE-Bench | 640.0 | Not measured | Only MiniMax M3 measured |
| LMArena Text: coding | 1470 | 1502 | Qwen3.8 Max (0902) |
| LMArena Text: software and IT | 1466 | 1498 | Qwen3.8 Max (0902) |
| LMArena WebDev: React | 1478 | 1692 | Qwen3.8 Max (0902) |
| LMArena WebDev: HTML | 1455 | 1649 | Qwen3.8 Max (0902) |
| LMArena WebDev: from an image | 1470 | 1637 | Qwen3.8 Max (0902) |
| Neither model has a result on the other 18 boards here | |||
IntelligenceLike-for-likePublic boards, shown as published, kept out of the index02
| Benchmark | MiniMax M3 | Qwen3.8 Max (0902) | Leader |
|---|---|---|---|
| AA Intelligence Index | 29.2 | 45.4 | Qwen3.8 Max (0902) |
| Epoch Capabilities Index | 146.9 | 157.4 | Qwen3.8 Max (0902) |
Agentic workLike-for-likePublic boards, shown as published, kept out of the index114
| Benchmark | MiniMax M3 | Qwen3.8 Max (0902) | Leader |
|---|---|---|---|
| AA Agentic Index | 29.5 | 56.0 | Qwen3.8 Max (0902) |
| Design Arena: agents building web apps | 1200 | 1300 | Qwen3.8 Max (0902) |
| Design Arena: agents building full-stack apps | 1197 | 1305 | Qwen3.8 Max (0902) |
| Design Arena: agents building mobile apps | 1169 | 1224 | Qwen3.8 Max (0902) |
| Design Arena: agents building Android apps | 1158 | 1190 | Qwen3.8 Max (0902) |
| Design Arena: agents building Godot games | Not measured | 1234 | Only Qwen3.8 Max (0902) measured |
| Design Arena: agents building games | 1159 | Not measured | Only MiniMax M3 measured |
| Design Arena: HTML slides | 1120 | Not measured | Only MiniMax M3 measured |
| Design Arena: PowerPoint slides | 1178 | 1232 | Qwen3.8 Max (0902) |
| Design Arena: spreadsheets | 1042 | 993 | MiniMax M3 |
| LiveBench research | 68.0% | 69.9% | Qwen3.8 Max (0902) |
| LiveBench browser | 8.0% | 32.0% | Qwen3.8 Max (0902) |
| LiveBench long agentic | 12.4% | 42.3% | Qwen3.8 Max (0902) |
| Finance Agent (Vals AI) | 48.3% | 50.6% | Qwen3.8 Max (0902) |
| Tax Agent Bench (Vals AI) | 20.6% | 32.1% | Qwen3.8 Max (0902) |
| Legal Research Bench (Vals AI) | 29.8% | 47.6% | Qwen3.8 Max (0902) |
| Excel Modeling Benchmark (Vals AI) | 47.8% | 60.1% | Qwen3.8 Max (0902) |
| Terminal-Bench-Science (Vals AI) | Not measured | 12.9% | Only Qwen3.8 Max (0902) measured |
| τ²-Bench Telecom | 88.9% | Not measured | Only MiniMax M3 measured |
| APEX-Agents | 37.7% | 63.3% | Qwen3.8 Max (0902) |
| OSWorld 2.0 | 4.6% | Not measured | Only MiniMax M3 measured |
| Vending-Bench 2 | $2,158 | Not measured | Only MiniMax M3 measured |
| LMArena Search | Not measured | 1261 | Only Qwen3.8 Max (0902) measured |
| Neither model has a result on the other 19 boards here | |||
ReasoningLike-for-likePublic boards, shown as published, kept out of the index112
| Benchmark | MiniMax M3 | Qwen3.8 Max (0902) | Leader |
|---|---|---|---|
| GPQA Diamond | 92.7% | 93.7% | Qwen3.8 Max (0902) |
| LiveBench reasoning | 74.5% | 88.2% | Qwen3.8 Max (0902) |
| MysteryMechanism (Vals AI) | Not measured | 23.9% | Only Qwen3.8 Max (0902) measured |
| MMMU Pro (Vals AI) | 81.2% | 88.0% | Qwen3.8 Max (0902) |
| GPQA Diamond (Artificial Analysis run) | 92.9% | 92.8% | MiniMax M3 |
| Humanity's Last Exam (Artificial Analysis run) | 39.0% | 43.1% | Qwen3.8 Max (0902) |
| GPQA Diamond (Epoch AI run) | 90.9% | 92.7% | Qwen3.8 Max (0902) |
| SimpleBench | 45.8% | Not measured | Only MiniMax M3 measured |
| Chess puzzles | 13.0% | 46.0% | Qwen3.8 Max (0902) |
| Mystery game puzzles | 8.0% | 38.0% | Qwen3.8 Max (0902) |
| LMArena Text: hard prompts | 1447 | 1496 | Qwen3.8 Max (0902) |
| LMArena Text: expert | 1461 | 1507 | Qwen3.8 Max (0902) |
| LMArena Text: hard prompts in English | 1455 | 1503 | Qwen3.8 Max (0902) |
| LMArena Vision: diagrams | 1266 | 1323 | Qwen3.8 Max (0902) |
| LMArena Vision: homework | 1268 | 1336 | Qwen3.8 Max (0902) |
| Neither model has a result on the other 8 boards here | |||
MathLike-for-likePublic boards, shown as published, kept out of the index04
| Benchmark | MiniMax M3 | Qwen3.8 Max (0902) | Leader |
|---|---|---|---|
| OTIS Mock AIME | 71.1% | 100.0% | Qwen3.8 Max (0902) |
| FrontierMath, tiers 1 to 3 | Not measured | 74.7% | Only Qwen3.8 Max (0902) measured |
| FrontierMath, tier 4 | Not measured | 46.3% | Only Qwen3.8 Max (0902) measured |
| ProofBench | 18.0% | 58.0% | Qwen3.8 Max (0902) |
| LMArena Text: math | 1429 | 1501 | Qwen3.8 Max (0902) |
| LMArena Text: mathematical work | 1448 | 1491 | Qwen3.8 Max (0902) |
| MathArena | Not measured | 51.6% | Only Qwen3.8 Max (0902) measured |
| Neither model has a result on the other 7 boards here | |||
KnowledgeLike-for-likePublic boards, shown as published, kept out of the index46
| Benchmark | MiniMax M3 | Qwen3.8 Max (0902) | Leader |
|---|---|---|---|
| MMLU Pro | 84.2% | 88.6% | Qwen3.8 Max (0902) |
| LegalBench (Vals AI) | 85.4% | 83.6% | MiniMax M3 |
| CorpFin (Vals AI) | 68.1% | 65.8% | MiniMax M3 |
| TaxEval (Vals AI) | 72.7% | 75.5% | Qwen3.8 Max (0902) |
| MortgageTax (Vals AI) | 68.4% | 64.0% | MiniMax M3 |
| MedCode (Vals AI) | 46.3% | 40.7% | MiniMax M3 |
| SimpleQA Verified | Not measured | 47.3% | Only Qwen3.8 Max (0902) measured |
| LMArena Text: science | 1452 | 1512 | Qwen3.8 Max (0902) |
| LMArena Text: medicine and healthcare | 1441 | 1503 | Qwen3.8 Max (0902) |
| LMArena Text: legal and government | 1437 | 1492 | Qwen3.8 Max (0902) |
| LMArena Text: business and finance | 1434 | 1480 | Qwen3.8 Max (0902) |
| Neither model has a result on the other 9 boards here | |||
Web designLike-for-likePublic boards, shown as published, kept out of the index08
| Benchmark | MiniMax M3 | Qwen3.8 Max (0902) | Leader |
|---|---|---|---|
| Design Arena: websites | 1257 | 1275 | Qwen3.8 Max (0902) |
| Design Arena: UI components | 1243 | 1336 | Qwen3.8 Max (0902) |
| Design Arena: game dev | 1210 | 1312 | Qwen3.8 Max (0902) |
| Design Arena: data viz | 1236 | 1282 | Qwen3.8 Max (0902) |
| Design Arena: 3D | 1208 | 1336 | Qwen3.8 Max (0902) |
| Design Arena: all code categories | 1246 | 1292 | Qwen3.8 Max (0902) |
| Design Arena: SVG | 1160 | Not measured | Only MiniMax M3 measured |
| Design Arena: ASCII art | 1161 | 1200 | Qwen3.8 Max (0902) |
| LiveBench design | 66.2% | 78.6% | Qwen3.8 Max (0902) |
GeneralLike-for-likePublic boards, shown as published, kept out of the index229
| Benchmark | MiniMax M3 | Qwen3.8 Max (0902) | Leader |
|---|---|---|---|
| LiveBench data analysis | 76.2% | 78.4% | Qwen3.8 Max (0902) |
| Public Benefits Bench (Vals AI) | 64.1% | 67.1% | Qwen3.8 Max (0902) |
| SAGE (Vals AI) | 50.6% | 51.2% | Qwen3.8 Max (0902) |
| MedScribe (Vals AI) | 87.3% | 85.0% | MiniMax M3 |
| Vals Multimodal Index | 60.0% | 65.4% | Qwen3.8 Max (0902) |
| IFBench | 82.9% | Not measured | Only MiniMax M3 measured |
| Long context reasoning (AA-LCR) | 83.0% | 80.3% | MiniMax M3 |
| DTBench | 78.9% | 92.0% | Qwen3.8 Max (0902) |
| LMCA | 33.7% | 46.2% | Qwen3.8 Max (0902) |
| Surface Evolver Bench | 55.0% | Not measured | Only MiniMax M3 measured |
| Furniture assembly | Not measured | 20.0% | Only Qwen3.8 Max (0902) measured |
| LMArena Text | 1433 | 1483 | Qwen3.8 Max (0902) |
| LMArena Text: instruction following | 1433 | 1481 | Qwen3.8 Max (0902) |
| LMArena Text: creative writing | 1404 | 1477 | Qwen3.8 Max (0902) |
| LMArena Vision | 1251 | 1313 | Qwen3.8 Max (0902) |
| LMArena Text: English | 1443 | 1491 | Qwen3.8 Max (0902) |
| LMArena Text: non-English | 1420 | 1473 | Qwen3.8 Max (0902) |
| LMArena Text: Chinese | 1466 | 1539 | Qwen3.8 Max (0902) |
| LMArena Text: German | 1428 | 1478 | Qwen3.8 Max (0902) |
| LMArena Text: Japanese | 1380 | 1473 | Qwen3.8 Max (0902) |
| LMArena Text: Korean | 1373 | 1460 | Qwen3.8 Max (0902) |
| LMArena Text: French | 1444 | 1498 | Qwen3.8 Max (0902) |
| LMArena Text: Spanish | 1433 | 1489 | Qwen3.8 Max (0902) |
| LMArena Text: Russian | 1426 | 1482 | Qwen3.8 Max (0902) |
| LMArena Text: Polish | 1423 | 1471 | Qwen3.8 Max (0902) |
| LMArena Text: multi-turn | 1443 | 1489 | Qwen3.8 Max (0902) |
| LMArena Text: longer queries | 1446 | 1488 | Qwen3.8 Max (0902) |
| LMArena Text: writing and language | 1419 | 1477 | Qwen3.8 Max (0902) |
| LMArena Text: entertainment and media | 1403 | 1464 | Qwen3.8 Max (0902) |
| LMArena Vision: English | 1263 | 1311 | Qwen3.8 Max (0902) |
| LMArena Vision: Chinese | 1311 | 1358 | Qwen3.8 Max (0902) |
| LMArena Vision: reading text | 1259 | 1322 | Qwen3.8 Max (0902) |
| LMArena Vision: humor | 1235 | 1304 | Qwen3.8 Max (0902) |
| LMArena Vision: recognizing things | 1197 | Not measured | Only MiniMax M3 measured |
| LMArena Vision: creative writing | 1234 | 1312 | Qwen3.8 Max (0902) |
| LMArena Document | 1435 | Not measured | Only MiniMax M3 measured |
| EQ-Bench 4 | 1150 | Not measured | Only MiniMax M3 measured |
| Neither model has a result on the other 11 boards here | |||
Image generationNot comparablePublic boards, shown as published, kept out of the index0 boards, none shared
| Benchmark | MiniMax M3 | Qwen3.8 Max (0902) | 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 M3 costs less
Repository review
50K input + 3K output tokens
MiniMax M3 costs less
Agent loop
200K input + 10K output tokens
MiniMax M3 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 | MiniMax M3 | Qwen3.8 Max (0902) |
|---|---|---|
| Maker | MiniMax | Qwen |
| Weights | Open weights | No public weights listed |
| Released | May 31, 2026 | September 3, 2026 |
| Context window | 1.05M | 1M |
| Input $/Mtok | $0.30 | $2.00 |
| Output $/Mtok | $1.20 | $6.00 |
| Coding index | 39.2 | 62.4 |
| Independent measurements behind it | 6 | 5 |
| Cost per solved task | $0.56 | $1.32 |
| Source | OpenRouter listing | OpenRouter listing |
Questions
Is MiniMax M3 better than Qwen3.8 Max (0902) for coding?
Pick Qwen3.8 Max (0902) for coding: it scores 62.4 on the Vibe Coding Index against 39.2 for MiniMax M3, though on 5 benchmarks against 6. Pick MiniMax M3 for output tokens 5.0x cheaper ($1.20 vs $6.00 per million) and a lower cost per solved task ($0.56 against $1.32).
Which is cheaper, MiniMax M3 or Qwen3.8 Max (0902)?
MiniMax M3 costs $0.30 per million input tokens and $1.20 per million output tokens; Qwen3.8 Max (0902) costs $2.00 and $6.00. Per solved SWE-bench Verified task, MiniMax M3 costs $0.56 and Qwen3.8 Max (0902) $1.32.
Which has the bigger context window, MiniMax M3 or Qwen3.8 Max (0902)?
MiniMax M3, with 1.05M tokens against 1M.
How many benchmarks is the MiniMax M3 vs Qwen3.8 Max (0902) comparison based on?
102 benchmarks have a result for both models as of October 22, 2026. MiniMax M3's coding index rests on 6 independent benchmarks and Qwen3.8 Max (0902)'s on 5. Where only one model has a result, the page says so instead of naming a winner.