GLM 5 vs MiniMax M2.7
Updated October 21, 2026. 60 benchmarks have a result for both. Coding is ranked on the Vibe Coding Index; every other board is shown as published.
Decision reading
Pick GLM 5 for coding: it scores 59.7 on the Vibe Coding Index against 28.4 for MiniMax M2.7.
Pick MiniMax M2.7 when output tokens cost 2.7x less, or when what matters is SWE-bench Verified (Vals AI) or GPQA Diamond, where it leads.
- Shared results
- 60
- GLM 5 only
- 15
- MiniMax M2.7 only
- 14
- Like-for-like categories
- 8 of 9
Shared results can support a head-to-head reading. Results only one model has describe coverage, not superiority.
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, four public coding benchmarks
GLM 5
GLM 5 scores 59.7 against 28.4.
2 and 1 benchmarks behind the scores
Intelligence
AA Intelligence Index
GLM 5
GLM 5 leads, 27.9 against 22.8.
Same board, both measured
Agentic work
AA Agentic Index
No call
Only MiniMax M2.7 has a AA Agentic Index result, 15.3.
Directional only
Reasoning
GPQA Diamond
MiniMax M2.7
MiniMax M2.7 leads, 86.6% against 83.3%.
Same board, both measured
Math
AIME
GLM 5
GLM 5 leads, 91.7% against 91.0%.
Same board, both measured
Knowledge
MMLU Pro
GLM 5
GLM 5 leads, 86.0% against 80.4%.
Same board, both measured
Web design
Design Arena: websites
GLM 5
GLM 5 leads, 1247 against 1243.
Same board, both measured
Long documents
Listed context window
No call
Both take 205K tokens.
Listed by OpenRouter
Everyday cost
1K input + 500 output tokens
MiniMax M2.7
GLM 5 $0.0026, MiniMax M2.7 $0.0009 per turn.
List prices
Cost per solved task
One SWE-bench Verified task, solved
GLM 5
GLM 5 $0.63, MiniMax M2.7 $0.64.
Published attempt cost and pass rate
At a glance
- Coding index
- GLM 5 59.7 from 2 benchmarks, MiniMax M2.7 28.4 from 1 benchmark.
- Widest category
- Agentic work, 13.5 points apart on average over 1 shared benchmark.
- Widest single gap
- LMArena Text: Japanese: GLM 5 1416 against MiniMax M2.7 1260, 26.0 points apart on a 0 to 100 scale.
- Output price
- GLM 5 $3.20 against MiniMax M2.7 $1.20 per million output tokens, 2.7x apart.
- Context window
- Both take 205K tokens.
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 | GLM 5 | Bars | MiniMax M2.7 | Ties |
|---|---|---|---|---|
| Coding | 11 | 1 | ||
| Intelligence | 2 | 0 | ||
| Agentic work | 1 | 0 | ||
| Reasoning | 3 | 3 | ||
| Math | 3 | 0 | ||
| Knowledge | 8 | 1 | ||
| Web design | 7 | 1 | ||
| General | 17 | 2 | ||
| All shared boards | 52 | 8 |
CodingLike-for-likeFour of these boards feed the Vibe Coding Index111
| Benchmark | GLM 5 | MiniMax M2.7 | Leader |
|---|---|---|---|
| SWE-bench Verified | 72.8% | Not measured | Only GLM 5 measured |
| SWE-bench Multilingual | 69.7% | Not measured | Only GLM 5 measured |
| LMArena WebDev | 1435 | 1399 | GLM 5 |
| AA Coding Index | Not measured | 52.6 | Only MiniMax M2.7 measured |
| LiveCodeBench (Vals AI) | 81.9% | 79.9% | GLM 5 |
| SWE-bench Verified (Vals AI) | 71.4% | 73.8% | MiniMax M2.7 |
| Vibe Code Bench (Vals AI) | 23.4% | 11.9% | GLM 5 |
| Code migration (Vals AI) | Not measured | 8.7% | Only MiniMax M2.7 measured |
| ProgramBench (Vals AI) | Not measured | 0.0% | Only MiniMax M2.7 measured |
| Terminal-Bench 2.1 (Vals AI) | Not measured | 48.7% | Only MiniMax M2.7 measured |
| SciCode | Not measured | 50.1% | Only MiniMax M2.7 measured |
| Terminal-Bench Hard (Artificial Analysis run) | 43.2% | 39.4% | GLM 5 |
| Terminal-Bench | 52.4% | 45.1% | GLM 5 |
| WeirdML | 48.2% | 37.0% | GLM 5 |
| ALE-Bench | 765.6 | 599.3 | GLM 5 |
| LMArena Text: coding | 1461 | 1454 | GLM 5 |
| LMArena Text: software and IT | 1460 | 1446 | GLM 5 |
| LMArena WebDev: React | 1424 | 1388 | GLM 5 |
| LMArena WebDev: HTML | 1450 | 1401 | GLM 5 |
| Neither model has a result on the other 19 boards here | |||
IntelligenceLike-for-likePublic boards, shown as published, kept out of the index20
| Benchmark | GLM 5 | MiniMax M2.7 | Leader |
|---|---|---|---|
| AA Intelligence Index | 27.9 | 22.8 | GLM 5 |
| Epoch Capabilities Index | 145.8 | 145.8 | GLM 5 |
Agentic workLike-for-likePublic boards, shown as published, kept out of the index10
| Benchmark | GLM 5 | MiniMax M2.7 | Leader |
|---|---|---|---|
| AA Agentic Index | Not measured | 15.3 | Only MiniMax M2.7 measured |
| Design Arena: agents building full-stack apps | 1123 | Not measured | Only GLM 5 measured |
| Design Arena: agents building mobile apps | 1136 | Not measured | Only GLM 5 measured |
| Design Arena: agents building Android apps | 1152 | Not measured | Only GLM 5 measured |
| Design Arena: agents building Godot games | 1144 | Not measured | Only GLM 5 measured |
| Design Arena: HTML slides | 1148 | Not measured | Only GLM 5 measured |
| Finance Agent (Vals AI) | Not measured | 27.9% | Only MiniMax M2.7 measured |
| Tax Agent Bench (Vals AI) | Not measured | 6.9% | Only MiniMax M2.7 measured |
| Legal Research Bench (Vals AI) | Not measured | 10.6% | Only MiniMax M2.7 measured |
| Excel Modeling Benchmark (Vals AI) | Not measured | 28.4% | Only MiniMax M2.7 measured |
| τ²-Bench Telecom | 98.2% | 84.8% | GLM 5 |
| Vending-Bench 2 | $4,432 | Not measured | Only GLM 5 measured |
| Neither model has a result on the other 27 boards here | |||
ReasoningLike-for-likePublic boards, shown as published, kept out of the index33
| Benchmark | GLM 5 | MiniMax M2.7 | Leader |
|---|---|---|---|
| GPQA Diamond | 83.3% | 86.6% | MiniMax M2.7 |
| GPQA Diamond (Artificial Analysis run) | 82.0% | 87.4% | MiniMax M2.7 |
| Humanity's Last Exam (Artificial Analysis run) | 29.3% | 29.6% | MiniMax M2.7 |
| GPQA Diamond (Epoch AI run) | 87.8% | Not measured | Only GLM 5 measured |
| ARC-AGI | 44.7% | Not measured | Only GLM 5 measured |
| ARC-AGI-2 | 4.9% | Not measured | Only GLM 5 measured |
| SimpleBench | 53.2% | Not measured | Only GLM 5 measured |
| LMArena Text: hard prompts | 1452 | 1422 | GLM 5 |
| LMArena Text: expert | 1454 | 1443 | GLM 5 |
| LMArena Text: hard prompts in English | 1460 | 1441 | GLM 5 |
| Neither model has a result on the other 11 boards here | |||
MathLike-for-likePublic boards, shown as published, kept out of the index30
| Benchmark | GLM 5 | MiniMax M2.7 | Leader |
|---|---|---|---|
| AIME | 91.7% | 91.0% | GLM 5 |
| OTIS Mock AIME | 80.0% | Not measured | Only GLM 5 measured |
| ProofBench | Not measured | 3.0% | Only MiniMax M2.7 measured |
| LMArena Text: math | 1440 | 1421 | GLM 5 |
| LMArena Text: mathematical work | 1451 | 1437 | GLM 5 |
| Neither model has a result on the other 8 boards here | |||
KnowledgeLike-for-likePublic boards, shown as published, kept out of the index81
| Benchmark | GLM 5 | MiniMax M2.7 | Leader |
|---|---|---|---|
| MMLU Pro | 86.0% | 80.4% | GLM 5 |
| LegalBench (Vals AI) | 84.1% | 84.0% | GLM 5 |
| CaseLaw (Vals AI) | 52.5% | 60.9% | MiniMax M2.7 |
| CorpFin (Vals AI) | 62.9% | 61.2% | GLM 5 |
| TaxEval (Vals AI) | 70.0% | 66.6% | GLM 5 |
| MedQA (Vals AI) | 94.3% | Not measured | Only GLM 5 measured |
| MedCode (Vals AI) | Not measured | 34.4% | Only MiniMax M2.7 measured |
| LMArena Text: science | 1460 | 1416 | GLM 5 |
| LMArena Text: medicine and healthcare | 1455 | 1409 | GLM 5 |
| LMArena Text: legal and government | 1453 | 1403 | GLM 5 |
| LMArena Text: business and finance | 1430 | 1407 | GLM 5 |
| Neither model has a result on the other 6 boards here | |||
Web designLike-for-likePublic boards, shown as published, kept out of the index71
| Benchmark | GLM 5 | MiniMax M2.7 | Leader |
|---|---|---|---|
| Design Arena: websites | 1247 | 1243 | GLM 5 |
| Design Arena: UI components | 1230 | 1210 | GLM 5 |
| Design Arena: game dev | 1234 | 1210 | GLM 5 |
| Design Arena: data viz | 1228 | 1233 | MiniMax M2.7 |
| Design Arena: 3D | 1229 | 1195 | GLM 5 |
| Design Arena: all code categories | 1246 | 1233 | GLM 5 |
| Design Arena: SVG | 1168 | 1136 | GLM 5 |
| Design Arena: ASCII art | 1146 | 1136 | GLM 5 |
| Neither model has a result on the other 1 board here | |||
GeneralLike-for-likePublic boards, shown as published, kept out of the index172
| Benchmark | GLM 5 | MiniMax M2.7 | Leader |
|---|---|---|---|
| MedScribe (Vals AI) | Not measured | 79.9% | Only MiniMax M2.7 measured |
| IFBench | 72.3% | 75.7% | MiniMax M2.7 |
| Long context reasoning (AA-LCR) | 75.7% | 78.3% | MiniMax M2.7 |
| CL-bench | 18.7% | Not measured | Only GLM 5 measured |
| ExploitBench | Not measured | 13.3% | Only MiniMax M2.7 measured |
| LMArena Text | 1446 | 1405 | GLM 5 |
| LMArena Text: instruction following | 1428 | 1405 | GLM 5 |
| LMArena Text: creative writing | 1438 | 1354 | GLM 5 |
| LMArena Text: English | 1458 | 1426 | GLM 5 |
| LMArena Text: non-English | 1430 | 1382 | GLM 5 |
| LMArena Text: Chinese | 1511 | 1442 | GLM 5 |
| LMArena Text: German | 1447 | 1398 | GLM 5 |
| LMArena Text: Japanese | 1416 | 1260 | GLM 5 |
| LMArena Text: Korean | 1423 | 1314 | GLM 5 |
| LMArena Text: French | 1456 | 1424 | GLM 5 |
| LMArena Text: Spanish | 1454 | 1402 | GLM 5 |
| LMArena Text: Russian | 1436 | 1383 | GLM 5 |
| LMArena Text: Polish | 1434 | 1387 | GLM 5 |
| LMArena Text: multi-turn | 1456 | 1412 | GLM 5 |
| LMArena Text: longer queries | 1446 | 1418 | GLM 5 |
| LMArena Text: writing and language | 1437 | 1376 | GLM 5 |
| LMArena Text: entertainment and media | 1426 | 1360 | GLM 5 |
| Neither model has a result on the other 24 boards here | |||
Image generationNot comparablePublic boards, shown as published, kept out of the index0 boards, none shared
| Benchmark | GLM 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 | Z.ai GLM 5 | MiniMax M2.7 |
|---|---|---|
| Maker | Z.ai | MiniMax |
| Weights | Open weights | Open weights |
| Released | February 11, 2026 | March 18, 2026 |
| Context window | 205K | 205K |
| Input $/Mtok | $1.00 | $0.30 |
| Output $/Mtok | $3.20 | $1.20 |
| Coding index | 59.7 | 28.4 |
| Independent measurements behind it | 2 | 1 |
| Cost per solved task | $0.63 | $0.64 |
| Source | OpenRouter listing | OpenRouter listing |
Questions
Is GLM 5 better than MiniMax M2.7 for coding?
Pick GLM 5 for coding: it scores 59.7 on the Vibe Coding Index against 28.4 for MiniMax M2.7. Pick MiniMax M2.7 when output tokens cost 2.7x less, or when what matters is SWE-bench Verified (Vals AI) or GPQA Diamond, where it leads.
Which is cheaper, GLM 5 or MiniMax M2.7?
GLM 5 costs $1.00 per million input tokens and $3.20 per million output tokens; MiniMax M2.7 costs $0.30 and $1.20. Per solved SWE-bench Verified task, GLM 5 costs $0.63 and MiniMax M2.7 $0.64.
Which has the bigger context window, GLM 5 or MiniMax M2.7?
Neither: both take 205K tokens.
How many benchmarks is the GLM 5 vs MiniMax M2.7 comparison based on?
60 benchmarks have a result for both models as of October 21, 2026. GLM 5's coding index rests on 2 independent benchmarks and MiniMax M2.7's on 1. Where only one model has a result, the page says so instead of naming a winner.