Model comparison
Codestral vs GLM-5.3
GLM-5.3 is the stronger model overall, scoring 54.8 to 30.6 on the Noometry Index. Codestral costs 4.8× less per token, which makes it the better buy when GLM-5.3's lead doesn't matter for your workload.
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 in 2 categories; 2 gaps are clear of the uncertainty.
- The widest gap is in coding, where GLM-5.3 leads 59.5 to 27.3.
- Codestral is cheaper at $0.30 / $0.90 per million input/output tokens, against $1.40 / $4.40 for GLM-5.3.
- GLM-5.3 accepts more context: 1M tokens versus 256K.
- GLM-5.3 has downloadable open weights; the other is API-only.
Side by side
| Codestral | GLM-5.3 | |
|---|---|---|
| Provider | Mistral AI | Z.ai (Zhipu) |
| Noometry Index | 30.6 | 54.8 |
| Released | 2024-05-29 | 2026-08-14 |
| Weights | Proprietary | Open |
| Context window | 256K | 1M |
| Max output | 8K | 131K |
| Input $ / M tokens | $0.30 | $1.40 |
| Output $ / M tokens | $0.90 | $4.40 |
| Results tracked | 7 | 42 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding GLM-5.3 leads
Codestral: 27.3 (#321), GLM-5.3: 59.5 (#14)
| Benchmark | Codestral | GLM-5.3 |
|---|---|---|
| ALE-Bench | 137.78 | 1,317 |
| DeepSWE | — | 69% |
| FrontierCode | — | 40.1% |
| Aider Polyglot | 11.1% | — |
| CursorBench | — | 42.6% |
| LMArena WebDev | — | 1622 |
| FrontierSWE | — | 30.2% |
| SciCode | — | 59% |
| WeirdML | — | 75.4% |
| BigCodeBench Instruct | 41.8% | — |
| LMArena Coding | — | 1496 |
| BigCodeBench Complete | 52.5% | — |
| HumanEval+ | 73.8% | — |
| MBPP+ | 61.9% | — |
Agentic & Tool Use Not comparable
Codestral: —, GLM-5.3: 36.4 (#38)
| Benchmark | Codestral | GLM-5.3 |
|---|---|---|
| APEX-Agents | — | 56.6% |
| Vending-Bench 2 | — | 8,164 |
Reasoning GLM-5.3 leads
Codestral: 19.8 (#251), GLM-5.3: 46.1 (#46)
| Benchmark | Codestral | GLM-5.3 |
|---|---|---|
| Kagi LLM Benchmark | 32.5% | — |
| NYT Connections (extended) | — | 74.2% |
| CritPt | — | 19.1% |
| Chess Puzzles | — | 21% |
| LMArena Hard Prompts | — | 1489 |
| Mystery Game Puzzles | — | 33% |
| DTBench | — | 87.7% |
| LMCA | — | 55.5% |
| Bench to the Future 3 | — | 0.15 |
| Epoch Capabilities Index | — | 155.61 |
Math Not comparable
Codestral: —, GLM-5.3: 62.3 (#33)
| Benchmark | Codestral | GLM-5.3 |
|---|---|---|
| FrontierMath (Tiers 1-3) | — | 68.8% |
| FrontierMath Tier 4 | — | 29.3% |
| OTIS Mock AIME 2024-2025 | — | 91.1% |
| ProofBench | — | 49% |
| LMArena Math | — | 1489 |
Knowledge Not comparable
Codestral: —, GLM-5.3: 58.3 (#37)
| Benchmark | Codestral | GLM-5.3 |
|---|---|---|
| GPQA Diamond | — | 90.9% |
| SimpleQA Verified | — | 41% |
| LMArena Expert | — | 1516 |
Multilingual Not comparable
Codestral: —, GLM-5.3: 55.7 (#28)
| Benchmark | Codestral | GLM-5.3 |
|---|---|---|
| LMArena Non-English | — | 1457 |
| LMArena Chinese | — | 1528 |
| LMArena French | — | 1499 |
| LMArena German | — | 1499 |
| LMArena Japanese | — | 1453 |
| LMArena Korean | — | 1472 |
| LMArena Russian | — | 1463 |
| LMArena Spanish | — | 1460 |
Instruction Following Not comparable
Codestral: —, GLM-5.3: 77.5 (#23)
| Benchmark | Codestral | GLM-5.3 |
|---|---|---|
| LMArena Instruction Following | — | 1477 |
Long Context Not comparable
Codestral: —, GLM-5.3: 45.4 (#41)
| Benchmark | Codestral | GLM-5.3 |
|---|---|---|
| LMArena Longer Query | — | 1482 |
Writing & Preference Not comparable
Codestral: —, GLM-5.3: 75.7 (#6)
| Benchmark | Codestral | GLM-5.3 |
|---|---|---|
| LMArena Text | — | 1471 |
| LMArena Creative Writing | — | 1457 |
| EQ-Bench Creative Writing | — | 2075 |
| LMArena Multi-Turn | — | 1472 |
Frequently asked questions
Is Codestral better than GLM-5.3?
GLM-5.3 is the stronger model overall, scoring 54.8 to 30.6 on the Noometry Index. Codestral costs 4.8× less per token, which makes it the better buy when GLM-5.3's lead doesn't matter for your workload.
Which is cheaper, Codestral or GLM-5.3?
Codestral is cheaper. It lists at $0.30 per million input tokens and $0.90 per million output tokens; GLM-5.3 lists at $1.40 and $4.40.
Is Codestral or GLM-5.3 better for coding?
GLM-5.3 scores higher on coding benchmarks: 59.5 versus 27.3 in the Noometry coding category.
Which has the bigger context window?
GLM-5.3 does, with 1M tokens against 256K.
How many benchmarks do Codestral and GLM-5.3 share?
1 benchmark has published results for both models. Codestral has 7 scored results on Noometry and GLM-5.3 has 42.