Model comparison
GLM-4.6 vs MiniMax-M3
MiniMax-M3 is the stronger model overall, scoring 43.8 to 41.4 on the Noometry Index.
Last verified . 21 shared benchmarks.
MiniMaxSummary
- They share 21 benchmarks with published results for both. GLM-4.6 scores higher in 2 categories and MiniMax-M3 in 7 categories; 5 gaps are clear of the uncertainty.
- The widest gap is in knowledge, where MiniMax-M3 leads 58.4 to 40.2.
- The biggest single-benchmark swing is SciCode: 38.4% for GLM-4.6 and 47.1% for MiniMax-M3.
- MiniMax-M3 is cheaper at $0.30 / $1.20 per million input/output tokens, against $0.60 / $2.20 for GLM-4.6.
- MiniMax-M3 accepts more context: 1M tokens versus 205K.
Side by side
| GLM-4.6 | MiniMax-M3 | |
|---|---|---|
| Provider | Z.ai (Zhipu) | MiniMax |
| Noometry Index | 41.4 | 43.8 |
| Released | 2025-09-30 | 2026-06-01 |
| Weights | Open | Open |
| Context window | 205K | 1M |
| Max output | 131K | 512K |
| Input $ / M tokens | $0.60 | $0.30 |
| Output $ / M tokens | $2.20 | $1.20 |
| Results tracked | 29 | 41 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding MiniMax-M3 leads
GLM-4.6: 40.1 (#148), MiniMax-M3: 41.8 (#118)
| Benchmark | GLM-4.6 | MiniMax-M3 |
|---|---|---|
| LMArena WebDev | 1340 | 1482 |
| SciCode | 38.4% | 47.1% |
| LMArena Coding | 1449 | 1469 |
| ALE-Bench | 340.82 | 640.02 |
| FrontierCode | — | 14.7% |
| SWE-bench Verified (bash only) | 55.4% | — |
Agentic & Tool Use GLM-4.6 leads
GLM-4.6: 32.3 (#66), MiniMax-M3: 22.6 (#130)
| Benchmark | GLM-4.6 | MiniMax-M3 |
|---|---|---|
| Terminal-Bench | 24.5% | — |
| APEX-Agents | — | 37.7% |
| Berkeley Function Calling Leaderboard | 72.4% | — |
| OSWorld 2.0 | — | 4.6% |
| GBAEval | — | 0.9% |
| Vending-Bench 2 | — | 2,158 |
Reasoning MiniMax-M3 leads
GLM-4.6: 23.7 (#172), MiniMax-M3: 30.1 (#87)
| Benchmark | GLM-4.6 | MiniMax-M3 |
|---|---|---|
| CritPt | 1.1% | 3.7% |
| LMArena Hard Prompts | 1440 | 1447 |
| SimpleBench | — | 45.8% |
| Kagi LLM Benchmark | 47.4% | — |
| NYT Connections (extended) | — | 65.1% |
| Chess Puzzles | — | 14% |
| Mystery Game Puzzles | — | 8% |
| DTBench | — | 78.9% |
| LMCA | — | 33.7% |
| Surface Evolver Bench | — | 55% |
| Epoch Capabilities Index | — | 146.95 |
| ForecastBench | — | 61.4 |
Math Too close to call
GLM-4.6: 39.1 (#111), MiniMax-M3: 40.0 (#95)
| Benchmark | GLM-4.6 | MiniMax-M3 |
|---|---|---|
| LMArena Math | 1432 | 1429 |
| OTIS Mock AIME 2024-2025 | — | 71.1% |
| ProofBench | — | 18% |
| FrontierMath (Feb 2025 set) | 3.8% | — |
| FrontierMath Tier 4 (v1) | 2.1% | — |
Knowledge MiniMax-M3 leads
GLM-4.6: 40.2 (#124), MiniMax-M3: 58.4 (#35)
| Benchmark | GLM-4.6 | MiniMax-M3 |
|---|---|---|
| LMArena Expert | 1431 | 1461 |
| GPQA Diamond | — | 90.9% |
| Vectara Hallucination Rate | 9.5% | — |
Multimodal Not comparable
GLM-4.6: —, MiniMax-M3: 40.2 (#51)
| Benchmark | GLM-4.6 | MiniMax-M3 |
|---|---|---|
| LMArena Vision | — | 1253 |
| LMArena Document | — | 1435 |
Multilingual Too close to call
GLM-4.6: 53.5 (#66), MiniMax-M3: 53.0 (#75)
| Benchmark | GLM-4.6 | MiniMax-M3 |
|---|---|---|
| LMArena Non-English | 1426 | 1420 |
| LMArena Chinese | 1499 | 1463 |
| LMArena French | 1459 | 1447 |
| LMArena German | 1447 | 1426 |
| LMArena Japanese | 1393 | 1381 |
| LMArena Korean | 1400 | 1372 |
| LMArena Russian | 1419 | 1428 |
| LMArena Spanish | 1436 | 1432 |
Instruction Following MiniMax-M3 leads
GLM-4.6: 74.3 (#98), MiniMax-M3: 75.5 (#62)
| Benchmark | GLM-4.6 | MiniMax-M3 |
|---|---|---|
| LMArena Instruction Following | 1410 | 1433 |
Long Context Too close to call
GLM-4.6: 43.4 (#94), MiniMax-M3: 44.2 (#72)
| Benchmark | GLM-4.6 | MiniMax-M3 |
|---|---|---|
| LMArena Longer Query | 1422 | 1445 |
Writing & Preference Too close to call
GLM-4.6: 61.1 (#90), MiniMax-M3: 62.1 (#83)
| Benchmark | GLM-4.6 | MiniMax-M3 |
|---|---|---|
| LMArena Text | 1440 | 1433 |
| LMArena Creative Writing | 1411 | 1404 |
| LMArena Multi-Turn | 1427 | 1442 |
| EQ-Bench Creative Writing | 1411 | — |
| EQ-Bench 4 | — | 1150 |
Frequently asked questions
Is GLM-4.6 better than MiniMax-M3?
MiniMax-M3 is the stronger model overall, scoring 43.8 to 41.4 on the Noometry Index.
Which is cheaper, GLM-4.6 or MiniMax-M3?
MiniMax-M3 is cheaper. It lists at $0.30 per million input tokens and $1.20 per million output tokens; GLM-4.6 lists at $0.60 and $2.20.
Is GLM-4.6 or MiniMax-M3 better for coding?
MiniMax-M3 scores higher on coding benchmarks: 41.8 versus 40.1 in the Noometry coding category.
Which has the bigger context window?
MiniMax-M3 does, with 1M tokens against 205K.
How many benchmarks do GLM-4.6 and MiniMax-M3 share?
21 benchmarks have published results for both models. GLM-4.6 has 29 scored results on Noometry and MiniMax-M3 has 41.