Model comparison
Codestral vs Qwen1.5-110B
Qwen1.5-110B is the stronger model overall, scoring 34.2 to 30.6 on the Noometry Index.
Last verified . 2 shared benchmarks.
Mistral AI
Alibaba (Qwen)Summary
- They share 2 benchmarks with published results for both. Codestral scores higher in 0 categories and Qwen1.5-110B in 2 categories; 2 gaps are clear of the uncertainty.
- The widest gap is in coding, where Qwen1.5-110B leads 33.0 to 27.3.
- The biggest single-benchmark swing is BigCodeBench Complete: 52.5% for Codestral and 44.4% for Qwen1.5-110B.
- Qwen1.5-110B has downloadable open weights; the other is API-only.
Side by side
| Codestral | Qwen1.5-110B | |
|---|---|---|
| Provider | Mistral AI | Alibaba (Qwen) |
| Noometry Index | 30.6 | 34.2 |
| Released | 2024-05-29 | 2024-04-25 |
| Weights | Proprietary | Open |
| Context window | 256K | — |
| Max output | 8K | — |
| Input $ / M tokens | $0.30 | — |
| Output $ / M tokens | $0.90 | — |
| Results tracked | 7 | 20 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Qwen1.5-110B leads
Codestral: 27.3 (#321), Qwen1.5-110B: 33.0 (#264)
| Benchmark | Codestral | Qwen1.5-110B |
|---|---|---|
| BigCodeBench Instruct | 41.8% | 35% |
| BigCodeBench Complete | 52.5% | 44.4% |
| Aider Polyglot | 11.1% | — |
| LMArena Coding | — | 1184 |
| ALE-Bench | 137.78 | — |
| HumanEval+ | 73.8% | — |
| MBPP+ | 61.9% | — |
Reasoning Qwen1.5-110B leads
Codestral: 19.8 (#251), Qwen1.5-110B: 22.7 (#189)
| Benchmark | Codestral | Qwen1.5-110B |
|---|---|---|
| Kagi LLM Benchmark | 32.5% | — |
| LMArena Hard Prompts | — | 1168 |
| ForecastBench | — | 57.7 |
Math Not comparable
Codestral: —, Qwen1.5-110B: 33.7 (#201)
| Benchmark | Codestral | Qwen1.5-110B |
|---|---|---|
| LMArena Math | — | 1185 |
Knowledge Not comparable
Codestral: —, Qwen1.5-110B: 31.2 (#219)
| Benchmark | Codestral | Qwen1.5-110B |
|---|---|---|
| LMArena Expert | — | 1144 |
Multilingual Not comparable
Codestral: —, Qwen1.5-110B: 33.6 (#250)
| Benchmark | Codestral | Qwen1.5-110B |
|---|---|---|
| LMArena Non-English | — | 1142 |
| LMArena Chinese | — | 1206 |
| LMArena French | — | 1151 |
| LMArena German | — | 1123 |
| LMArena Japanese | — | 1074 |
| LMArena Korean | — | 1044 |
| LMArena Russian | — | 1118 |
| LMArena Spanish | — | 1142 |
Instruction Following Not comparable
Codestral: —, Qwen1.5-110B: 60.3 (#252)
| Benchmark | Codestral | Qwen1.5-110B |
|---|---|---|
| LMArena Instruction Following | — | 1158 |
Long Context Not comparable
Codestral: —, Qwen1.5-110B: 35.1 (#242)
| Benchmark | Codestral | Qwen1.5-110B |
|---|---|---|
| LMArena Longer Query | — | 1157 |
Writing & Preference Not comparable
Codestral: —, Qwen1.5-110B: 38.0 (#255)
| Benchmark | Codestral | Qwen1.5-110B |
|---|---|---|
| LMArena Text | — | 1175 |
| LMArena Creative Writing | — | 1148 |
| LMArena Multi-Turn | — | 1160 |
Frequently asked questions
Is Codestral better than Qwen1.5-110B?
Qwen1.5-110B is the stronger model overall, scoring 34.2 to 30.6 on the Noometry Index.
Is Codestral or Qwen1.5-110B better for coding?
Qwen1.5-110B scores higher on coding benchmarks: 33.0 versus 27.3 in the Noometry coding category.
How many benchmarks do Codestral and Qwen1.5-110B share?
2 benchmarks have published results for both models. Codestral has 7 scored results on Noometry and Qwen1.5-110B has 20.