Model comparison
Codestral vs GPT-4 Turbo
Codestral and GPT-4 Turbo score almost the same on the Noometry Index (30.6 vs 30.5), so choose on price, context window or the category you care about most.
Last verified . 4 shared benchmarks.
Mistral AISummary
- They share 4 benchmarks with published results for both. Codestral scores higher in 1 category and GPT-4 Turbo in 1 category; 2 gaps are clear of the uncertainty.
- The widest gap is in coding, where GPT-4 Turbo leads 33.8 to 27.3.
- The biggest single-benchmark swing is BigCodeBench Instruct: 41.8% for Codestral and 48.2% for GPT-4 Turbo.
- Codestral is cheaper at $0.30 / $0.90 per million input/output tokens, against $10 / $30 for GPT-4 Turbo.
- Codestral accepts more context: 256K tokens versus 128K.
Side by side
| Codestral | GPT-4 Turbo | |
|---|---|---|
| Provider | Mistral AI | OpenAI |
| Noometry Index | 30.6 | 30.5 |
| Released | 2024-05-29 | 2023-11-06 |
| Weights | Proprietary | Proprietary |
| Context window | 256K | 128K |
| Max output | 8K | 4K |
| Input $ / M tokens | $0.30 | $10 |
| Output $ / M tokens | $0.90 | $30 |
| Results tracked | 7 | 36 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding GPT-4 Turbo leads
Codestral: 27.3 (#321), GPT-4 Turbo: 33.8 (#249)
| Benchmark | Codestral | GPT-4 Turbo |
|---|---|---|
| BigCodeBench Instruct | 41.8% | 48.2% |
| BigCodeBench Complete | 52.5% | 58.2% |
| HumanEval+ | 73.8% | 86.6% |
| MBPP+ | 61.9% | 73.3% |
| Aider Polyglot | 11.1% | — |
| WeirdML | — | 18% |
| LMArena Coding | — | 1268 |
| ALE-Bench | 137.78 | — |
Agentic & Tool Use Not comparable
Codestral: —, GPT-4 Turbo: —
| Benchmark | Codestral | GPT-4 Turbo |
|---|---|---|
| METR Time Horizons | — | 36.7% |
Reasoning Codestral leads
Codestral: 19.8 (#251), GPT-4 Turbo: 15.3 (#317)
| Benchmark | Codestral | GPT-4 Turbo |
|---|---|---|
| SimpleBench | — | 25.1% |
| Kagi LLM Benchmark | 32.5% | — |
| Chess Puzzles | — | 6% |
| LMArena Hard Prompts | — | 1251 |
| DTBench | — | 61.6% |
| LMCA | — | 9.8% |
| Epoch Capabilities Index | — | 127.25 |
| ForecastBench | — | 59.4 |
Math Not comparable
Codestral: —, GPT-4 Turbo: 9.0 (#322)
| Benchmark | Codestral | GPT-4 Turbo |
|---|---|---|
| FrontierMath (Tiers 1-3) | — | 0.7% |
| OTIS Mock AIME 2024-2025 | — | 6.7% |
| LMArena Math | — | 1272 |
| MATH Level 5 | — | 46.7% |
Knowledge Not comparable
Codestral: —, GPT-4 Turbo: 24.3 (#268)
| Benchmark | Codestral | GPT-4 Turbo |
|---|---|---|
| GPQA Diamond | — | 46.6% |
| Confabulations | — | 28.4% |
| LMArena Expert | — | 1223 |
| MMLU | — | 81.3% |
Multimodal Not comparable
Codestral: —, GPT-4 Turbo: 30.6 (#110)
| Benchmark | Codestral | GPT-4 Turbo |
|---|---|---|
| LMArena Vision | — | 1090 |
Multilingual Not comparable
Codestral: —, GPT-4 Turbo: 40.5 (#216)
| Benchmark | Codestral | GPT-4 Turbo |
|---|---|---|
| LMArena Non-English | — | 1245 |
| LMArena Chinese | — | 1242 |
| LMArena French | — | 1276 |
| LMArena German | — | 1259 |
| LMArena Japanese | — | 1194 |
| LMArena Korean | — | 1187 |
| LMArena Russian | — | 1259 |
| LMArena Spanish | — | 1260 |
Instruction Following Not comparable
Codestral: —, GPT-4 Turbo: 65.8 (#216)
| Benchmark | Codestral | GPT-4 Turbo |
|---|---|---|
| LMArena Instruction Following | — | 1249 |
Long Context Not comparable
Codestral: —, GPT-4 Turbo: 38.0 (#206)
| Benchmark | Codestral | GPT-4 Turbo |
|---|---|---|
| LMArena Longer Query | — | 1254 |
Writing & Preference Not comparable
Codestral: —, GPT-4 Turbo: 47.7 (#206)
| Benchmark | Codestral | GPT-4 Turbo |
|---|---|---|
| LMArena Text | — | 1272 |
| LMArena Creative Writing | — | 1269 |
| LMArena Multi-Turn | — | 1267 |
Frequently asked questions
Is Codestral better than GPT-4 Turbo?
Codestral and GPT-4 Turbo score almost the same on the Noometry Index (30.6 vs 30.5), so choose on price, context window or the category you care about most.
Which is cheaper, Codestral or GPT-4 Turbo?
Codestral is cheaper. It lists at $0.30 per million input tokens and $0.90 per million output tokens; GPT-4 Turbo lists at $10 and $30.
Is Codestral or GPT-4 Turbo better for coding?
GPT-4 Turbo scores higher on coding benchmarks: 33.8 versus 27.3 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-4 Turbo share?
4 benchmarks have published results for both models. Codestral has 7 scored results on Noometry and GPT-4 Turbo has 36.