Model comparison
Codestral vs Inkling
Inkling is the stronger model overall, scoring 44.1 to 30.6 on the Noometry Index. Codestral costs 5.7× less per token, which makes it the better buy when Inkling's lead doesn't matter for your workload.
Last verified . 1 shared benchmarks.
Mistral AISummary
- They share 1 benchmark with published results for both. Codestral scores higher in 0 categories and Inkling in 2 categories; 2 gaps are clear of the uncertainty.
- The widest gap is in reasoning, where Inkling leads 40.4 to 19.8.
- Codestral is cheaper at $0.30 / $0.90 per million input/output tokens, against $1.87 / $4.68 for Inkling.
- Codestral accepts more context: 256K tokens versus 66K.
- Inkling has downloadable open weights; the other is API-only.
Side by side
| Codestral | Inkling | |
|---|---|---|
| Provider | Mistral AI | Thinking Machines Lab |
| Noometry Index | 30.6 | 44.1 |
| Released | 2024-05-29 | 2026-07-15 |
| Weights | Proprietary | Open |
| Context window | 256K | 66K |
| Max output | 8K | 66K |
| Input $ / M tokens | $0.30 | $1.87 |
| Output $ / M tokens | $0.90 | $4.68 |
| Results tracked | 7 | 41 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Inkling leads
Codestral: 27.3 (#321), Inkling: 34.5 (#234)
| Benchmark | Codestral | Inkling |
|---|---|---|
| ALE-Bench | 137.78 | 946 |
| FrontierCode | — | 14% |
| Aider Polyglot | 11.1% | — |
| LMArena WebDev | — | 1413 |
| FrontierSWE | — | 4.1% |
| SciCode | — | 47% |
| WeirdML | — | 32.3% |
| BigCodeBench Instruct | 41.8% | — |
| LMArena Coding | — | 1464 |
| BigCodeBench Complete | 52.5% | — |
| HumanEval+ | 73.8% | — |
| MBPP+ | 61.9% | — |
Agentic & Tool Use Not comparable
Codestral: —, Inkling: 29.6 (#85)
| Benchmark | Codestral | Inkling |
|---|---|---|
| APEX-Agents | — | 33.8% |
| τ²-bench Banking | — | 25% |
Reasoning Inkling leads
Codestral: 19.8 (#251), Inkling: 40.4 (#56)
| Benchmark | Codestral | Inkling |
|---|---|---|
| ARC-AGI-2 | — | 36.5% |
| SimpleBench | — | 50% |
| Kagi LLM Benchmark | 32.5% | — |
| ARC-AGI-1 | — | 79.5% |
| CritPt | — | 5.4% |
| Chess Puzzles | — | 21% |
| LMArena Hard Prompts | — | 1451 |
| DTBench | — | 87.5% |
| LMCA | — | 37.6% |
| Epoch Capabilities Index | — | 148.54 |
Math Not comparable
Codestral: —, Inkling: 31.3 (#225)
| Benchmark | Codestral | Inkling |
|---|---|---|
| FrontierMath (Tiers 1-3) | — | 33.3% |
| FrontierMath Tier 4 | — | 4.9% |
| OTIS Mock AIME 2024-2025 | — | 88.9% |
| ProofBench | — | 0% |
| LMArena Math | — | 1479 |
Knowledge Not comparable
Codestral: —, Inkling: 55.1 (#49)
| Benchmark | Codestral | Inkling |
|---|---|---|
| GPQA Diamond | — | 88.3% |
| SimpleQA Verified | — | 40.3% |
| LMArena Expert | — | 1465 |
Multilingual Not comparable
Codestral: —, Inkling: 54.0 (#52)
| Benchmark | Codestral | Inkling |
|---|---|---|
| LMArena Non-English | — | 1434 |
| LMArena Chinese | — | 1490 |
| LMArena French | — | 1458 |
| LMArena German | — | 1446 |
| LMArena Japanese | — | 1429 |
| LMArena Korean | — | 1404 |
| LMArena Russian | — | 1429 |
| LMArena Spanish | — | 1448 |
Instruction Following Not comparable
Codestral: —, Inkling: 75.1 (#71)
| Benchmark | Codestral | Inkling |
|---|---|---|
| LMArena Instruction Following | — | 1426 |
Long Context Not comparable
Codestral: —, Inkling: 43.8 (#86)
| Benchmark | Codestral | Inkling |
|---|---|---|
| LMArena Longer Query | — | 1434 |
Writing & Preference Not comparable
Codestral: —, Inkling: 65.2 (#51)
| Benchmark | Codestral | Inkling |
|---|---|---|
| LMArena Text | — | 1441 |
| LMArena Creative Writing | — | 1387 |
| EQ-Bench Creative Writing | — | 1611 |
| EQ-Bench 4 | — | 1226 |
| LMArena Multi-Turn | — | 1436 |
Frequently asked questions
Is Codestral better than Inkling?
Inkling is the stronger model overall, scoring 44.1 to 30.6 on the Noometry Index. Codestral costs 5.7× less per token, which makes it the better buy when Inkling's lead doesn't matter for your workload.
Which is cheaper, Codestral or Inkling?
Codestral is cheaper. It lists at $0.30 per million input tokens and $0.90 per million output tokens; Inkling lists at $1.87 and $4.68.
Is Codestral or Inkling better for coding?
Inkling scores higher on coding benchmarks: 34.5 versus 27.3 in the Noometry coding category.
Which has the bigger context window?
Codestral does, with 256K tokens against 66K.
How many benchmarks do Codestral and Inkling share?
1 benchmark has published results for both models. Codestral has 7 scored results on Noometry and Inkling has 41.