Model comparison
GLM-5.3-Flash vs MiniMax-M2.5
GLM-5.3-Flash is the stronger model overall, scoring 51.8 to 38.3 on the Noometry Index.
Last verified . 23 shared benchmarks.
MiniMaxSummary
- They share 23 benchmarks with published results for both. GLM-5.3-Flash scores higher in 9 categories and MiniMax-M2.5 in 0 categories; 9 gaps are clear of the uncertainty.
- The widest gap is in reasoning, where GLM-5.3-Flash leads 48.0 to 17.5.
- The biggest single-benchmark swing is ARC-AGI-2: 65.8% for GLM-5.3-Flash and 4.9% for MiniMax-M2.5.
- GLM-5.3-Flash is cheaper at $0.15 / $0.50 per million input/output tokens, against $0.30 / $1.20 for MiniMax-M2.5.
- GLM-5.3-Flash accepts more context: 1M tokens versus 205K.
Side by side
| GLM-5.3-Flash | MiniMax-M2.5 | |
|---|---|---|
| Provider | Z.ai (Zhipu) | MiniMax |
| Noometry Index | 51.8 | 38.3 |
| Released | 2026-08-20 | 2026-02-12 |
| Weights | Open | Open |
| Context window | 1M | 205K |
| Max output | 131K | 131K |
| Input $ / M tokens | $0.15 | $0.30 |
| Output $ / M tokens | $0.50 | $1.20 |
| Results tracked | 40 | 33 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding GLM-5.3-Flash leads
GLM-5.3-Flash: 53.1 (#31), MiniMax-M2.5: 48.1 (#58)
| Benchmark | GLM-5.3-Flash | MiniMax-M2.5 |
|---|---|---|
| LMArena WebDev | 1609 | 1387 |
| LMArena Coding | 1508 | 1381 |
| ALE-Bench | 303.55 | 618.17 |
| DeepSWE | 63.4% | — |
| FrontierCode | 31.8% | — |
| SWE-bench Verified (bash only) | — | 75.8% |
| CursorBench | 36.8% | — |
| SWE-bench Multilingual | — | 68.3% |
| FrontierSWE | 18.1% | — |
| SciCode | 51.6% | — |
Agentic & Tool Use GLM-5.3-Flash leads
GLM-5.3-Flash: 34.2 (#47), MiniMax-M2.5: 30.4 (#77)
| Benchmark | GLM-5.3-Flash | MiniMax-M2.5 |
|---|---|---|
| Terminal-Bench | — | 42.7% |
| APEX-Agents | 52.8% | — |
| GDP.pdf | 14% | — |
| Vending-Bench 2 | — | -23.16 |
Reasoning GLM-5.3-Flash leads
GLM-5.3-Flash: 48.0 (#42), MiniMax-M2.5: 17.5 (#292)
| Benchmark | GLM-5.3-Flash | MiniMax-M2.5 |
|---|---|---|
| ARC-AGI-2 | 65.8% | 4.9% |
| ARC-AGI-1 | 91% | 63.7% |
| LMArena Hard Prompts | 1491 | 1372 |
| Epoch Capabilities Index | 151.88 | 146.68 |
| Kagi LLM Benchmark | — | 55.2% |
| NYT Connections (extended) | — | 16.8% |
| CritPt | 15.4% | — |
| Chess Puzzles | 14% | — |
| Mystery Game Puzzles | 8% | — |
| Surface Evolver Bench | 52.5% | — |
| Bench to the Future 3 | 0.15 | — |
Math GLM-5.3-Flash leads
GLM-5.3-Flash: 53.3 (#47), MiniMax-M2.5: 26.9 (#253)
| Benchmark | GLM-5.3-Flash | MiniMax-M2.5 |
|---|---|---|
| ProofBench | 21% | 4% |
| LMArena Math | 1500 | 1378 |
| FrontierMath (Tiers 1-3) | 55.8% | — |
| FrontierMath Tier 4 | 17.1% | — |
| OTIS Mock AIME 2024-2025 | 93.9% | — |
Knowledge GLM-5.3-Flash leads
GLM-5.3-Flash: 58.4 (#36), MiniMax-M2.5: 39.2 (#135)
| Benchmark | GLM-5.3-Flash | MiniMax-M2.5 |
|---|---|---|
| LMArena Expert | 1513 | 1379 |
| GPQA Diamond | 90.2% | — |
| Vectara Hallucination Rate | — | 9.1% |
Multimodal Not comparable
GLM-5.3-Flash: 42.8 (#27), MiniMax-M2.5: —
| Benchmark | GLM-5.3-Flash | MiniMax-M2.5 |
|---|---|---|
| LMArena Vision | 1296 | — |
Multilingual GLM-5.3-Flash leads
GLM-5.3-Flash: 56.0 (#25), MiniMax-M2.5: 47.1 (#152)
| Benchmark | GLM-5.3-Flash | MiniMax-M2.5 |
|---|---|---|
| LMArena Non-English | 1462 | 1338 |
| LMArena Chinese | 1527 | 1393 |
| LMArena French | 1496 | 1362 |
| LMArena German | 1470 | 1362 |
| LMArena Japanese | 1429 | 1171 |
| LMArena Korean | 1446 | 1232 |
| LMArena Russian | 1469 | 1358 |
| LMArena Spanish | 1471 | 1354 |
Instruction Following GLM-5.3-Flash leads
GLM-5.3-Flash: 77.5 (#20), MiniMax-M2.5: 71.5 (#148)
| Benchmark | GLM-5.3-Flash | MiniMax-M2.5 |
|---|---|---|
| LMArena Instruction Following | 1478 | 1353 |
Long Context GLM-5.3-Flash leads
GLM-5.3-Flash: 45.4 (#39), MiniMax-M2.5: 37.5 (#216)
| Benchmark | GLM-5.3-Flash | MiniMax-M2.5 |
|---|---|---|
| LMArena Longer Query | 1482 | 1366 |
| CL-bench | — | 11.4% |
| CL-bench Life | — | 6.3% |
Writing & Preference GLM-5.3-Flash leads
GLM-5.3-Flash: 65.3 (#50), MiniMax-M2.5: 53.9 (#153)
| Benchmark | GLM-5.3-Flash | MiniMax-M2.5 |
|---|---|---|
| LMArena Text | 1471 | 1359 |
| LMArena Creative Writing | 1442 | 1331 |
| LMArena Multi-Turn | 1467 | 1364 |
| EQ-Bench Creative Writing | — | 1361 |
Frequently asked questions
Is GLM-5.3-Flash better than MiniMax-M2.5?
GLM-5.3-Flash is the stronger model overall, scoring 51.8 to 38.3 on the Noometry Index.
Which is cheaper, GLM-5.3-Flash or MiniMax-M2.5?
GLM-5.3-Flash is cheaper. It lists at $0.15 per million input tokens and $0.50 per million output tokens; MiniMax-M2.5 lists at $0.30 and $1.20.
Is GLM-5.3-Flash or MiniMax-M2.5 better for coding?
GLM-5.3-Flash scores higher on coding benchmarks: 53.1 versus 48.1 in the Noometry coding category.
Which has the bigger context window?
GLM-5.3-Flash does, with 1M tokens against 205K.
How many benchmarks do GLM-5.3-Flash and MiniMax-M2.5 share?
23 benchmarks have published results for both models. GLM-5.3-Flash has 40 scored results on Noometry and MiniMax-M2.5 has 33.