Model comparison
Codestral vs Gemma 3 12B
Gemma 3 12B is the stronger model overall, scoring 31.8 to 30.1 on the Noometry Index.
Last verified . 0 shared benchmarks.
Mistral AI
GoogleSummary
- The widest gap is in coding, where Gemma 3 12B leads 31.2 to 26.9.
- Gemma 3 12B is cheaper at $0.05 / $0.15 per million input/output tokens, against $0.30 / $0.90 for Codestral.
- Codestral accepts more context: 256K tokens versus 131K.
Side by side
| Codestral | Gemma 3 12B | |
|---|---|---|
| Provider | Mistral AI | |
| Noometry Index | 30.1 | 31.8 |
| Released | 2024-05-29 | 2025-03-12 |
| Weights | Open | Open |
| Context window | 256K | 131K |
| Max output | 4K | 8K |
| Input $ / M tokens | $0.30 | $0.05 |
| Output $ / M tokens | $0.90 | $0.15 |
| Results tracked | 7 | 24 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Gemma 3 12B leads
Codestral: 26.9 (#316), Gemma 3 12B: 31.2 (#276)
| Benchmark | Codestral | Gemma 3 12B |
|---|---|---|
| Aider Polyglot | 11.1% | — |
| SciCode | — | 17.4% |
| BigCodeBench Instruct | 41.8% | — |
| LMArena Coding | — | 1281 |
| BigCodeBench Complete | 52.5% | — |
| ALE-Bench | 137.78 | — |
| HumanEval+ | 73.8% | — |
| MBPP+ | 61.9% | — |
Agentic & Tool Use Not comparable
Codestral: —, Gemma 3 12B: 25.3 (#101)
| Benchmark | Codestral | Gemma 3 12B |
|---|---|---|
| Berkeley Function Calling Leaderboard | — | 30.4% |
Reasoning Codestral leads
Codestral: 19.1 (#244), Gemma 3 12B: 15.2 (#307)
| Benchmark | Codestral | Gemma 3 12B |
|---|---|---|
| Kagi LLM Benchmark | 32.5% | — |
| CritPt | — | 0% |
| Chess Puzzles | — | 0% |
| LMArena Hard Prompts | — | 1309 |
| DTBench | — | 48.8% |
| LMCA | — | 4.5% |
| Epoch Capabilities Index | — | 123.49 |
Math Not comparable
Codestral: —, Gemma 3 12B: 21.0 (#273)
| Benchmark | Codestral | Gemma 3 12B |
|---|---|---|
| OTIS Mock AIME 2024-2025 | — | 16.7% |
| LMArena Math | — | 1307 |
Knowledge Not comparable
Codestral: —, Gemma 3 12B: 26.5 (#250)
| Benchmark | Codestral | Gemma 3 12B |
|---|---|---|
| GPQA Diamond | — | 39.5% |
| Vectara Hallucination Rate | — | 4.4% |
| LMArena Expert | — | 1248 |
Multimodal Not comparable
Codestral: —, Gemma 3 12B: —
| Benchmark | Codestral | Gemma 3 12B |
|---|---|---|
| MindCube | — | 46.7% |
Multilingual Not comparable
Codestral: —, Gemma 3 12B: 45.9 (#160)
| Benchmark | Codestral | Gemma 3 12B |
|---|---|---|
| LMArena Non-English | — | 1318 |
| LMArena German | — | 1370 |
| LMArena Russian | — | 1335 |
Instruction Following Not comparable
Codestral: —, Gemma 3 12B: 68.8 (#179)
| Benchmark | Codestral | Gemma 3 12B |
|---|---|---|
| LMArena Instruction Following | — | 1299 |
Long Context Not comparable
Codestral: —, Gemma 3 12B: 40.0 (#157)
| Benchmark | Codestral | Gemma 3 12B |
|---|---|---|
| LMArena Longer Query | — | 1317 |
Writing & Preference Not comparable
Codestral: —, Gemma 3 12B: 47.7 (#205)
| Benchmark | Codestral | Gemma 3 12B |
|---|---|---|
| LMArena Text | — | 1334 |
| LMArena Creative Writing | — | 1331 |
| EQ-Bench Creative Writing | — | 1126 |
| LMArena Multi-Turn | — | 1334 |
Frequently asked questions
Is Codestral better than Gemma 3 12B?
Gemma 3 12B is the stronger model overall, scoring 31.8 to 30.1 on the Noometry Index.
Which is cheaper, Codestral or Gemma 3 12B?
Gemma 3 12B is cheaper. It lists at $0.05 per million input tokens and $0.15 per million output tokens; Codestral lists at $0.30 and $0.90.
Is Codestral or Gemma 3 12B better for coding?
Gemma 3 12B scores higher on coding benchmarks: 31.2 versus 26.9 in the Noometry coding category.
Which has the bigger context window?
Codestral does, with 256K tokens against 131K.
How many benchmarks do Codestral and Gemma 3 12B share?
0 benchmarks have published results for both models. Codestral has 7 scored results on Noometry and Gemma 3 12B has 24.