Model comparison
Codestral vs Qwen3 14B
Qwen3 14B is the stronger model overall, scoring 35.5 to 30.6 on the Noometry Index.
Last verified . 1 shared benchmarks.
Mistral AI
Alibaba (Qwen)Summary
- They share 1 benchmark with published results for both. Codestral scores higher in 1 category and Qwen3 14B in 1 category; 2 gaps are clear of the uncertainty.
- The widest gap is in coding, where Qwen3 14B leads 37.3 to 27.3.
- The biggest single-benchmark swing is Kagi LLM Benchmark: 32.5% for Codestral and 49.1% for Qwen3 14B.
- Codestral is cheaper at $0.30 / $0.90 per million input/output tokens, against $0.35 / $1.40 for Qwen3 14B.
- Codestral accepts more context: 256K tokens versus 131K.
- Qwen3 14B has downloadable open weights; the other is API-only.
Side by side
| Codestral | Qwen3 14B | |
|---|---|---|
| Provider | Mistral AI | Alibaba (Qwen) |
| Noometry Index | 30.6 | 35.5 |
| Released | 2024-05-29 | 2025-04 |
| Weights | Proprietary | Open |
| Context window | 256K | 131K |
| Max output | 8K | 8K |
| Input $ / M tokens | $0.30 | $0.35 |
| Output $ / M tokens | $0.90 | $1.40 |
| Results tracked | 7 | 12 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Qwen3 14B leads
Codestral: 27.3 (#321), Qwen3 14B: 37.3 (#195)
| Benchmark | Codestral | Qwen3 14B |
|---|---|---|
| Aider Polyglot | 11.1% | — |
| SciCode | — | 31.6% |
| BigCodeBench Instruct | 41.8% | — |
| BigCodeBench Complete | 52.5% | — |
| ALE-Bench | 137.78 | — |
| HumanEval+ | 73.8% | — |
| MBPP+ | 61.9% | — |
Agentic & Tool Use Not comparable
Codestral: —, Qwen3 14B: 29.6 (#83)
| Benchmark | Codestral | Qwen3 14B |
|---|---|---|
| Berkeley Function Calling Leaderboard | — | 41% |
Reasoning Codestral leads
Codestral: 19.8 (#251), Qwen3 14B: 18.5 (#280)
| Benchmark | Codestral | Qwen3 14B |
|---|---|---|
| Kagi LLM Benchmark | 32.5% | 49.1% |
| CritPt | — | 0% |
| Chess Puzzles | — | 4% |
| DTBench | — | 64% |
| LMCA | — | 18.2% |
| Epoch Capabilities Index | — | 138.23 |
Math Not comparable
Codestral: —, Qwen3 14B: 38.6 (#133)
| Benchmark | Codestral | Qwen3 14B |
|---|---|---|
| OTIS Mock AIME 2024-2025 | — | 66.4% |
Knowledge Not comparable
Codestral: —, Qwen3 14B: 39.3 (#134)
| Benchmark | Codestral | Qwen3 14B |
|---|---|---|
| GPQA Diamond | — | 63.8% |
| Vectara Hallucination Rate | — | 5.4% |
Long Context Not comparable
Codestral: —, Qwen3 14B: 38.1 (#204)
| Benchmark | Codestral | Qwen3 14B |
|---|---|---|
| Fiction.LiveBench | — | 62.5% |
Frequently asked questions
Is Codestral better than Qwen3 14B?
Qwen3 14B is the stronger model overall, scoring 35.5 to 30.6 on the Noometry Index.
Which is cheaper, Codestral or Qwen3 14B?
Codestral is cheaper. It lists at $0.30 per million input tokens and $0.90 per million output tokens; Qwen3 14B lists at $0.35 and $1.40.
Is Codestral or Qwen3 14B better for coding?
Qwen3 14B scores higher on coding benchmarks: 37.3 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 Qwen3 14B share?
1 benchmark has published results for both models. Codestral has 7 scored results on Noometry and Qwen3 14B has 12.