Model comparison
Codestral vs DeepSeek-V3.2-Exp
DeepSeek-V3.2-Exp is the stronger model overall, scoring 44.3 to 30.6 on the Noometry Index.
Last verified . 2 shared benchmarks.
Mistral AI
DeepSeekSummary
- They share 2 benchmarks with published results for both. Codestral scores higher in 0 categories and DeepSeek-V3.2-Exp in 2 categories; 2 gaps are clear of the uncertainty.
- The widest gap is in coding, where DeepSeek-V3.2-Exp leads 46.5 to 27.3.
- The biggest single-benchmark swing is Aider Polyglot: 11.1% for Codestral and 74.2% for DeepSeek-V3.2-Exp.
- DeepSeek-V3.2-Exp is cheaper at $0.26 / $0.38 per million input/output tokens, against $0.30 / $0.90 for Codestral.
- Codestral accepts more context: 256K tokens versus 164K.
- DeepSeek-V3.2-Exp has downloadable open weights; the other is API-only.
Side by side
| Codestral | DeepSeek-V3.2-Exp | |
|---|---|---|
| Provider | Mistral AI | DeepSeek |
| Noometry Index | 30.6 | 44.3 |
| Released | 2024-05-29 | 2025-09-29 |
| Weights | Proprietary | Open |
| Context window | 256K | 164K |
| Max output | 8K | 66K |
| Input $ / M tokens | $0.30 | $0.26 |
| Output $ / M tokens | $0.90 | $0.38 |
| Results tracked | 7 | 49 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding DeepSeek-V3.2-Exp leads
Codestral: 27.3 (#321), DeepSeek-V3.2-Exp: 46.5 (#65)
| Benchmark | Codestral | DeepSeek-V3.2-Exp |
|---|---|---|
| Aider Polyglot | 11.1% | 74.2% |
| SWE-bench Verified (bash only) | — | 70% |
| LMArena WebDev | — | 1362 |
| SWE-bench Multilingual | — | 59% |
| SciCode | — | 38.9% |
| WeirdML | — | 39.5% |
| BigCodeBench Instruct | 41.8% | — |
| LMArena Coding | — | 1454 |
| BigCodeBench Complete | 52.5% | — |
| ALE-Bench | 137.78 | — |
| HumanEval+ | 73.8% | — |
| MBPP+ | 61.9% | — |
Agentic & Tool Use Not comparable
Codestral: —, DeepSeek-V3.2-Exp: 32.7 (#59)
| Benchmark | Codestral | DeepSeek-V3.2-Exp |
|---|---|---|
| Terminal-Bench | — | 39.6% |
| APEX-Agents | — | 21.3% |
| Berkeley Function Calling Leaderboard | — | 56.7% |
| TheAgentCompany | — | 42.9% |
| Vending-Bench 2 | — | 1,034 |
Reasoning DeepSeek-V3.2-Exp leads
Codestral: 19.8 (#251), DeepSeek-V3.2-Exp: 22.1 (#208)
| Benchmark | Codestral | DeepSeek-V3.2-Exp |
|---|---|---|
| Kagi LLM Benchmark | 32.5% | 52.2% |
| ARC-AGI-2 | — | 4% |
| NYT Connections (extended) | — | 36.7% |
| ARC-AGI-1 | — | 57% |
| CritPt | — | 2.9% |
| Chess Puzzles | — | 14% |
| Thematic Generalization | — | 65% |
| LMArena Hard Prompts | — | 1434 |
| DTBench | — | 87.7% |
| LMCA | — | 29.1% |
| Epoch Capabilities Index | — | 146.27 |
Math Not comparable
Codestral: —, DeepSeek-V3.2-Exp: 41.7 (#87)
| Benchmark | Codestral | DeepSeek-V3.2-Exp |
|---|---|---|
| MathArena Final-Answer Competitions | — | 57.7% |
| OTIS Mock AIME 2024-2025 | — | 87.8% |
| ProofBench | — | 8% |
| LMArena Math | — | 1435 |
| FrontierMath (Feb 2025 set) | — | 22.1% |
| FrontierMath Tier 4 (v1) | — | 2.1% |
Knowledge Not comparable
Codestral: —, DeepSeek-V3.2-Exp: 51.7 (#66)
| Benchmark | Codestral | DeepSeek-V3.2-Exp |
|---|---|---|
| GPQA Diamond | — | 83.4% |
| Vectara Hallucination Rate | — | 5.3% |
| LMArena Expert | — | 1436 |
Multilingual Not comparable
Codestral: —, DeepSeek-V3.2-Exp: 52.2 (#90)
| Benchmark | Codestral | DeepSeek-V3.2-Exp |
|---|---|---|
| LMArena Non-English | — | 1409 |
| LMArena Chinese | — | 1461 |
| LMArena French | — | 1433 |
| LMArena German | — | 1440 |
| LMArena Japanese | — | 1374 |
| LMArena Korean | — | 1371 |
| LMArena Russian | — | 1424 |
| LMArena Spanish | — | 1440 |
Instruction Following Not comparable
Codestral: —, DeepSeek-V3.2-Exp: 74.5 (#93)
| Benchmark | Codestral | DeepSeek-V3.2-Exp |
|---|---|---|
| LMArena Instruction Following | — | 1413 |
Long Context Not comparable
Codestral: —, DeepSeek-V3.2-Exp: 47.6 (#16)
| Benchmark | Codestral | DeepSeek-V3.2-Exp |
|---|---|---|
| Fiction.LiveBench | — | 83.3% |
| CL-bench | — | 13.2% |
| CL-bench Life | — | 9.5% |
| LMArena Longer Query | — | 1428 |
Writing & Preference Not comparable
Codestral: —, DeepSeek-V3.2-Exp: 62.4 (#77)
| Benchmark | Codestral | DeepSeek-V3.2-Exp |
|---|---|---|
| LMArena Text | — | 1425 |
| LMArena Creative Writing | — | 1403 |
| EQ-Bench Creative Writing | — | 1515 |
| LMArena Multi-Turn | — | 1427 |
Frequently asked questions
Is Codestral better than DeepSeek-V3.2-Exp?
DeepSeek-V3.2-Exp is the stronger model overall, scoring 44.3 to 30.6 on the Noometry Index.
Which is cheaper, Codestral or DeepSeek-V3.2-Exp?
DeepSeek-V3.2-Exp is cheaper. It lists at $0.26 per million input tokens and $0.38 per million output tokens; Codestral lists at $0.30 and $0.90.
Is Codestral or DeepSeek-V3.2-Exp better for coding?
DeepSeek-V3.2-Exp scores higher on coding benchmarks: 46.5 versus 27.3 in the Noometry coding category.
Which has the bigger context window?
Codestral does, with 256K tokens against 164K.
How many benchmarks do Codestral and DeepSeek-V3.2-Exp share?
2 benchmarks have published results for both models. Codestral has 7 scored results on Noometry and DeepSeek-V3.2-Exp has 49.