Model comparison
Codestral vs GLM-5.3-Flash
GLM-5.3-Flash is the stronger model overall, scoring 51.8 to 30.6 on the Noometry Index.
Last verified . 1 shared benchmarks.
Mistral AISummary
- They share 1 benchmark with published results for both. Codestral scores higher in 0 categories and GLM-5.3-Flash in 2 categories; 2 gaps are clear of the uncertainty.
- The widest gap is in reasoning, where GLM-5.3-Flash leads 48.0 to 19.8.
- GLM-5.3-Flash is cheaper at $0.15 / $0.50 per million input/output tokens, against $0.30 / $0.90 for Codestral.
- GLM-5.3-Flash accepts more context: 1M tokens versus 256K.
- GLM-5.3-Flash has downloadable open weights; the other is API-only.
Side by side
| Codestral | GLM-5.3-Flash | |
|---|---|---|
| Provider | Mistral AI | Z.ai (Zhipu) |
| Noometry Index | 30.6 | 51.8 |
| Released | 2024-05-29 | 2026-08-20 |
| Weights | Proprietary | Open |
| Context window | 256K | 1M |
| Max output | 8K | 131K |
| Input $ / M tokens | $0.30 | $0.15 |
| Output $ / M tokens | $0.90 | $0.50 |
| Results tracked | 7 | 40 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding GLM-5.3-Flash leads
Codestral: 27.3 (#321), GLM-5.3-Flash: 53.1 (#31)
| Benchmark | Codestral | GLM-5.3-Flash |
|---|---|---|
| ALE-Bench | 137.78 | 303.55 |
| DeepSWE | — | 63.4% |
| FrontierCode | — | 31.8% |
| Aider Polyglot | 11.1% | — |
| CursorBench | — | 36.8% |
| LMArena WebDev | — | 1609 |
| FrontierSWE | — | 18.1% |
| SciCode | — | 51.6% |
| BigCodeBench Instruct | 41.8% | — |
| LMArena Coding | — | 1508 |
| BigCodeBench Complete | 52.5% | — |
| HumanEval+ | 73.8% | — |
| MBPP+ | 61.9% | — |
Agentic & Tool Use Not comparable
Codestral: —, GLM-5.3-Flash: 34.2 (#47)
| Benchmark | Codestral | GLM-5.3-Flash |
|---|---|---|
| APEX-Agents | — | 52.8% |
| GDP.pdf | — | 14% |
Reasoning GLM-5.3-Flash leads
Codestral: 19.8 (#251), GLM-5.3-Flash: 48.0 (#42)
| Benchmark | Codestral | GLM-5.3-Flash |
|---|---|---|
| ARC-AGI-2 | — | 65.8% |
| Kagi LLM Benchmark | 32.5% | — |
| ARC-AGI-1 | — | 91% |
| CritPt | — | 15.4% |
| Chess Puzzles | — | 14% |
| LMArena Hard Prompts | — | 1491 |
| Mystery Game Puzzles | — | 8% |
| Surface Evolver Bench | — | 52.5% |
| Bench to the Future 3 | — | 0.15 |
| Epoch Capabilities Index | — | 151.88 |
Math Not comparable
Codestral: —, GLM-5.3-Flash: 53.3 (#47)
| Benchmark | Codestral | GLM-5.3-Flash |
|---|---|---|
| FrontierMath (Tiers 1-3) | — | 55.8% |
| FrontierMath Tier 4 | — | 17.1% |
| OTIS Mock AIME 2024-2025 | — | 93.9% |
| ProofBench | — | 21% |
| LMArena Math | — | 1500 |
Knowledge Not comparable
Codestral: —, GLM-5.3-Flash: 58.4 (#36)
| Benchmark | Codestral | GLM-5.3-Flash |
|---|---|---|
| GPQA Diamond | — | 90.2% |
| LMArena Expert | — | 1513 |
Multimodal Not comparable
Codestral: —, GLM-5.3-Flash: 42.8 (#27)
| Benchmark | Codestral | GLM-5.3-Flash |
|---|---|---|
| LMArena Vision | — | 1296 |
Multilingual Not comparable
Codestral: —, GLM-5.3-Flash: 56.0 (#25)
| Benchmark | Codestral | GLM-5.3-Flash |
|---|---|---|
| LMArena Non-English | — | 1462 |
| LMArena Chinese | — | 1527 |
| LMArena French | — | 1496 |
| LMArena German | — | 1470 |
| LMArena Japanese | — | 1429 |
| LMArena Korean | — | 1446 |
| LMArena Russian | — | 1469 |
| LMArena Spanish | — | 1471 |
Instruction Following Not comparable
Codestral: —, GLM-5.3-Flash: 77.5 (#20)
| Benchmark | Codestral | GLM-5.3-Flash |
|---|---|---|
| LMArena Instruction Following | — | 1478 |
Long Context Not comparable
Codestral: —, GLM-5.3-Flash: 45.4 (#39)
| Benchmark | Codestral | GLM-5.3-Flash |
|---|---|---|
| LMArena Longer Query | — | 1482 |
Writing & Preference Not comparable
Codestral: —, GLM-5.3-Flash: 65.3 (#50)
| Benchmark | Codestral | GLM-5.3-Flash |
|---|---|---|
| LMArena Text | — | 1471 |
| LMArena Creative Writing | — | 1442 |
| LMArena Multi-Turn | — | 1467 |
Frequently asked questions
Is Codestral better than GLM-5.3-Flash?
GLM-5.3-Flash is the stronger model overall, scoring 51.8 to 30.6 on the Noometry Index.
Which is cheaper, Codestral or GLM-5.3-Flash?
GLM-5.3-Flash is cheaper. It lists at $0.15 per million input tokens and $0.50 per million output tokens; Codestral lists at $0.30 and $0.90.
Is Codestral or GLM-5.3-Flash better for coding?
GLM-5.3-Flash scores higher on coding benchmarks: 53.1 versus 27.3 in the Noometry coding category.
Which has the bigger context window?
GLM-5.3-Flash does, with 1M tokens against 256K.
How many benchmarks do Codestral and GLM-5.3-Flash share?
1 benchmark has published results for both models. Codestral has 7 scored results on Noometry and GLM-5.3-Flash has 40.