Model comparison
Codestral vs gpt-oss-20b
gpt-oss-20b is the stronger model overall, scoring 32.5 to 30.6 on the Noometry Index.
Last verified . 2 shared benchmarks.
Mistral AISummary
- They share 2 benchmarks with published results for both. Codestral scores higher in 1 category and gpt-oss-20b in 1 category; one gap is clear of the uncertainty.
- The widest gap is in coding, where gpt-oss-20b leads 37.6 to 27.3.
- The biggest single-benchmark swing is Kagi LLM Benchmark: 32.5% for Codestral and 53.2% for gpt-oss-20b.
- gpt-oss-20b is cheaper at $0.018 / $0.09 per million input/output tokens, against $0.30 / $0.90 for Codestral.
- Codestral accepts more context: 256K tokens versus 131K.
- gpt-oss-20b has downloadable open weights; the other is API-only.
Side by side
| Codestral | gpt-oss-20b | |
|---|---|---|
| Provider | Mistral AI | OpenAI |
| Noometry Index | 30.6 | 32.5 |
| Released | 2024-05-29 | 2025-08-05 |
| Weights | Proprietary | Open |
| Context window | 256K | 131K |
| Max output | 8K | 16K |
| Input $ / M tokens | $0.30 | $0.018 |
| Output $ / M tokens | $0.90 | $0.09 |
| Results tracked | 7 | 34 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding gpt-oss-20b leads
Codestral: 27.3 (#321), gpt-oss-20b: 37.6 (#192)
| Benchmark | Codestral | gpt-oss-20b |
|---|---|---|
| ALE-Bench | 137.78 | 566.05 |
| Aider Polyglot | 11.1% | — |
| SciCode | — | 34.4% |
| WeirdML | — | 40.9% |
| BigCodeBench Instruct | 41.8% | — |
| LMArena Coding | — | 1306 |
| BigCodeBench Complete | 52.5% | — |
| HumanEval+ | 73.8% | — |
| MBPP+ | 61.9% | — |
Agentic & Tool Use Not comparable
Codestral: —, gpt-oss-20b: 9.3 (#154)
| Benchmark | Codestral | gpt-oss-20b |
|---|---|---|
| Terminal-Bench | — | 3.4% |
Reasoning Too close to call
Codestral: 19.8 (#251), gpt-oss-20b: 19.3 (#261)
| Benchmark | Codestral | gpt-oss-20b |
|---|---|---|
| Kagi LLM Benchmark | 32.5% | 53.2% |
| CritPt | — | 1.4% |
| Chess Puzzles | — | 4% |
| LMArena Hard Prompts | — | 1274 |
| DTBench | — | 68% |
| LMCA | — | 14.5% |
| Epoch Capabilities Index | — | 137.82 |
Math Not comparable
Codestral: —, gpt-oss-20b: 39.4 (#103)
| Benchmark | Codestral | gpt-oss-20b |
|---|---|---|
| OTIS Mock AIME 2024-2025 | — | 65.3% |
| Omni-MATH | — | 56.5% |
| LMArena Math | — | 1317 |
Knowledge Not comparable
Codestral: —, gpt-oss-20b: 34.6 (#195)
| Benchmark | Codestral | gpt-oss-20b |
|---|---|---|
| GPQA Diamond | — | 60.8% |
| MMLU-Pro | — | 74% |
| GPQA (HELM) | — | 59.4% |
| LMArena Expert | — | 1258 |
Multilingual Not comparable
Codestral: —, gpt-oss-20b: 42.2 (#197)
| Benchmark | Codestral | gpt-oss-20b |
|---|---|---|
| LMArena Non-English | — | 1268 |
| LMArena Chinese | — | 1314 |
| LMArena German | — | 1255 |
| LMArena Japanese | — | 1244 |
| LMArena Korean | — | 1236 |
| LMArena Russian | — | 1278 |
| LMArena Spanish | — | 1267 |
Instruction Following Not comparable
Codestral: —, gpt-oss-20b: 61.8 (#240)
| Benchmark | Codestral | gpt-oss-20b |
|---|---|---|
| IFEval | — | 73.2% |
| LMArena Instruction Following | — | 1236 |
Long Context Not comparable
Codestral: —, gpt-oss-20b: 37.9 (#209)
| Benchmark | Codestral | gpt-oss-20b |
|---|---|---|
| LMArena Longer Query | — | 1250 |
Writing & Preference Not comparable
Codestral: —, gpt-oss-20b: 35.5 (#265)
| Benchmark | Codestral | gpt-oss-20b |
|---|---|---|
| LMArena Text | — | 1287 |
| LMArena Creative Writing | — | 1201 |
| EQ-Bench Creative Writing | — | 666 |
| WildBench | — | 73.7% |
| LMArena Multi-Turn | — | 1268 |
Frequently asked questions
Is Codestral better than gpt-oss-20b?
gpt-oss-20b is the stronger model overall, scoring 32.5 to 30.6 on the Noometry Index.
Which is cheaper, Codestral or gpt-oss-20b?
gpt-oss-20b is cheaper. It lists at $0.018 per million input tokens and $0.09 per million output tokens; Codestral lists at $0.30 and $0.90.
Is Codestral or gpt-oss-20b better for coding?
gpt-oss-20b scores higher on coding benchmarks: 37.6 versus 27.3 in the Noometry coding category.
Which has the bigger context window?
Codestral does, with 256K tokens against 131K.
How many benchmarks do Codestral and gpt-oss-20b share?
2 benchmarks have published results for both models. Codestral has 7 scored results on Noometry and gpt-oss-20b has 34.