Model comparison
Codestral vs GPT-4o mini
Codestral is the stronger model overall, scoring 30.6 to 25.5 on the Noometry Index. GPT-4o mini costs 1.7× less per token, which makes it the better buy when Codestral's lead doesn't matter for your workload.
Last verified . 6 shared benchmarks.
Mistral AISummary
- They share 6 benchmarks with published results for both. Codestral scores higher in 2 categories and GPT-4o mini in 0 categories; 2 gaps are clear of the uncertainty.
- The widest gap is in reasoning, where Codestral leads 19.8 to 8.7.
- The biggest single-benchmark swing is Aider Polyglot: 11.1% for Codestral and 3.6% for GPT-4o mini.
- GPT-4o mini is cheaper at $0.15 / $0.60 per million input/output tokens, against $0.30 / $0.90 for Codestral.
- Codestral accepts more context: 256K tokens versus 128K.
Side by side
| Codestral | GPT-4o mini | |
|---|---|---|
| Provider | Mistral AI | OpenAI |
| Noometry Index | 30.6 | 25.5 |
| Released | 2024-05-29 | 2024-07-18 |
| Weights | Proprietary | Proprietary |
| Context window | 256K | 128K |
| Max output | 8K | 16K |
| Input $ / M tokens | $0.30 | $0.15 |
| Output $ / M tokens | $0.90 | $0.60 |
| Results tracked | 7 | 60 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Codestral leads
Codestral: 27.3 (#321), GPT-4o mini: 22.0 (#335)
| Benchmark | Codestral | GPT-4o mini |
|---|---|---|
| Aider Polyglot | 11.1% | 3.6% |
| BigCodeBench Instruct | 41.8% | 46.1% |
| BigCodeBench Complete | 52.5% | 57.4% |
| HumanEval+ | 73.8% | 83.5% |
| MBPP+ | 61.9% | 72.2% |
| WeirdML | — | 11.8% |
| LiveBench Coding | — | 43.1% |
| LMArena Coding | — | 1290 |
| ALE-Bench | 137.78 | — |
Agentic & Tool Use Not comparable
Codestral: —, GPT-4o mini: 27.5 (#101)
| Benchmark | Codestral | GPT-4o mini |
|---|---|---|
| BALROG | — | 17.4% |
Reasoning Codestral leads
Codestral: 19.8 (#251), GPT-4o mini: 8.7 (#347)
| Benchmark | Codestral | GPT-4o mini |
|---|---|---|
| Kagi LLM Benchmark | 32.5% | 28.8% |
| ARC-AGI-2 | — | 0% |
| SimpleBench | — | 10.7% |
| Chess Puzzles | — | 0% |
| LiveBench Reasoning | — | 32.8% |
| LMArena Hard Prompts | — | 1267 |
| Mystery Game Puzzles | — | 12% |
| DTBench | — | 54.4% |
| LiveBench Data Analysis | — | 50% |
| LMCA | — | 10.4% |
| Epoch Capabilities Index | — | 126.56 |
| LiveBench | — | 41.3% |
| PIQA | — | 88.7% |
Math Not comparable
Codestral: —, GPT-4o mini: 10.4 (#314)
| Benchmark | Codestral | GPT-4o mini |
|---|---|---|
| FrontierMath (Tiers 1-3) | — | 0.7% |
| OTIS Mock AIME 2024-2025 | — | 6.9% |
| Omni-MATH | — | 28% |
| LiveBench Math | — | 36.3% |
| LMArena Math | — | 1267 |
| MATH Level 5 | — | 52.6% |
| GSM8K | — | 91.3% |
Knowledge Not comparable
Codestral: —, GPT-4o mini: 17.7 (#284)
| Benchmark | Codestral | GPT-4o mini |
|---|---|---|
| GPQA Diamond | — | 37.7% |
| SimpleQA Verified | — | 8.3% |
| MMLU-Pro | — | 60.3% |
| Confabulations | — | 37.2% |
| GPQA (HELM) | — | 36.8% |
| LMArena Expert | — | 1235 |
| BoolQ | — | 88.7% |
| MMLU | — | 81.8% |
Multimodal Not comparable
Codestral: —, GPT-4o mini: 25.9 (#122)
| Benchmark | Codestral | GPT-4o mini |
|---|---|---|
| LMArena Vision | — | 1066 |
| Video-MME | — | 64.8% |
| GeoBench | — | 64% |
| VPCT | — | 34% |
Multilingual Not comparable
Codestral: —, GPT-4o mini: 42.0 (#199)
| Benchmark | Codestral | GPT-4o mini |
|---|---|---|
| LMArena Non-English | — | 1266 |
| LMArena Chinese | — | 1265 |
| LMArena French | — | 1297 |
| LMArena German | — | 1272 |
| LMArena Japanese | — | 1216 |
| LMArena Korean | — | 1195 |
| LMArena Russian | — | 1275 |
| LMArena Spanish | — | 1276 |
Instruction Following Not comparable
Codestral: —, GPT-4o mini: 61.9 (#239)
| Benchmark | Codestral | GPT-4o mini |
|---|---|---|
| LiveBench Instruction Following | — | 56.8% |
| IFEval | — | 78.2% |
| LMArena Instruction Following | — | 1258 |
Long Context Not comparable
Codestral: —, GPT-4o mini: 39.1 (#186)
| Benchmark | Codestral | GPT-4o mini |
|---|---|---|
| LMArena Longer Query | — | 1289 |
Writing & Preference Not comparable
Codestral: —, GPT-4o mini: 39.5 (#248)
| Benchmark | Codestral | GPT-4o mini |
|---|---|---|
| LMArena Text | — | 1286 |
| LMArena Creative Writing | — | 1268 |
| Short-Story Creative Writing | — | 67.2% |
| EQ-Bench Creative Writing | — | 873 |
| WildBench | — | 79.1% |
| LMArena Multi-Turn | — | 1285 |
| LiveBench Language | — | 28.6% |
Frequently asked questions
Is Codestral better than GPT-4o mini?
Codestral is the stronger model overall, scoring 30.6 to 25.5 on the Noometry Index. GPT-4o mini costs 1.7× less per token, which makes it the better buy when Codestral's lead doesn't matter for your workload.
Which is cheaper, Codestral or GPT-4o mini?
GPT-4o mini is cheaper. It lists at $0.15 per million input tokens and $0.60 per million output tokens; Codestral lists at $0.30 and $0.90.
Is Codestral or GPT-4o mini better for coding?
Codestral scores higher on coding benchmarks: 27.3 versus 22.0 in the Noometry coding category.
Which has the bigger context window?
Codestral does, with 256K tokens against 128K.
How many benchmarks do Codestral and GPT-4o mini share?
6 benchmarks have published results for both models. Codestral has 7 scored results on Noometry and GPT-4o mini has 60.