Model comparison
Devstral Small 2505 vs Grok 4.5
Grok 4.5 is the stronger model overall, scoring 55.0 to 34.3 on the Noometry Index. Devstral Small 2505 costs 20× less per token, which makes it the better buy when Grok 4.5's lead doesn't matter for your workload.
Last verified . 3 shared benchmarks.
Mistral AISummary
- They share 3 benchmarks with published results for both. Devstral Small 2505 scores higher in 0 categories and Grok 4.5 in 2 categories; 2 gaps are clear of the uncertainty.
- The widest gap is in reasoning, where Grok 4.5 leads 56.1 to 19.7.
- The biggest single-benchmark swing is Kagi LLM Benchmark: 37.7% for Devstral Small 2505 and 83.5% for Grok 4.5.
- Devstral Small 2505 is cheaper at $0.10 / $0.30 per million input/output tokens, against $2 / $6 for Grok 4.5.
- Grok 4.5 accepts more context: 500K tokens versus 128K.
- Devstral Small 2505 has downloadable open weights; the other is API-only.
Side by side
| Devstral Small 2505 | Grok 4.5 | |
|---|---|---|
| Provider | Mistral AI | xAI |
| Noometry Index | 34.3 | 55.0 |
| Released | 2025-05-07 | 2026-07-08 |
| Weights | Open | Proprietary |
| Context window | 128K | 500K |
| Max output | 128K | 500K |
| Input $ / M tokens | $0.10 | $2 |
| Output $ / M tokens | $0.30 | $6 |
| Results tracked | 4 | 52 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Grok 4.5 leads
Devstral Small 2505: 38.9 (#166), Grok 4.5: 52.2 (#35)
| Benchmark | Devstral Small 2505 | Grok 4.5 |
|---|---|---|
| SciCode | 28.8% | 54.1% |
| DeepSWE | — | 53.8% |
| FrontierCode | — | 42.4% |
| SWE-bench Verified (bash only) | 56.4% | — |
| LMArena WebDev | — | 1553 |
| WeirdML | — | 46.4% |
| LMArena Coding | — | 1474 |
| ALE-Bench | — | 1,309 |
Agentic & Tool Use Not comparable
Devstral Small 2505: —, Grok 4.5: 44.4 (#17)
| Benchmark | Devstral Small 2505 | Grok 4.5 |
|---|---|---|
| APEX-Agents | — | 56.2% |
| τ²-bench Banking | — | 47.9% |
| PostTrainBench | — | 23.4% |
| GBAEval | — | 65.4% |
| GDP.pdf | — | 14% |
| LMArena Search | — | 1213 |
| Vending-Bench 2 | — | 3,887 |
Reasoning Grok 4.5 leads
Devstral Small 2505: 19.7 (#252), Grok 4.5: 56.1 (#25)
| Benchmark | Devstral Small 2505 | Grok 4.5 |
|---|---|---|
| Kagi LLM Benchmark | 37.7% | 83.5% |
| CritPt | 0% | 15.4% |
| ARC-AGI-2 | — | 52.6% |
| SimpleBench | — | 70% |
| NYT Connections (extended) | — | 79.9% |
| ARC-AGI-1 | — | 87.2% |
| Chess Puzzles | — | 36% |
| LMArena Hard Prompts | — | 1462 |
| DTBench | — | 96.5% |
| LMCA | — | 45.2% |
| Surface Evolver Bench | — | 74.4% |
| Epoch Capabilities Index | — | 153.92 |
Math Not comparable
Devstral Small 2505: —, Grok 4.5: 60.9 (#35)
| Benchmark | Devstral Small 2505 | Grok 4.5 |
|---|---|---|
| FrontierMath (Tiers 1-3) | — | 57.2% |
| FrontierMath Tier 4 | — | 24.4% |
| OTIS Mock AIME 2024-2025 | — | 97.8% |
| ProofBench | — | 31% |
| LMArena Math | — | 1459 |
Knowledge Not comparable
Devstral Small 2505: —, Grok 4.5: 62.3 (#24)
| Benchmark | Devstral Small 2505 | Grok 4.5 |
|---|---|---|
| GPQA Diamond | — | 93.4% |
| SimpleQA Verified | — | 48.3% |
| LMArena Expert | — | 1466 |
Multimodal Not comparable
Devstral Small 2505: —, Grok 4.5: 37.6 (#72)
| Benchmark | Devstral Small 2505 | Grok 4.5 |
|---|---|---|
| LMArena Vision | — | 1288 |
| Blueprint-Bench 2 | — | 27.3% |
| Furniture Assembly | — | 22.5% |
| LMArena Document | — | 1452 |
Multilingual Not comparable
Devstral Small 2505: —, Grok 4.5: 54.4 (#42)
| Benchmark | Devstral Small 2505 | Grok 4.5 |
|---|---|---|
| LMArena Non-English | — | 1440 |
| LMArena Chinese | — | 1496 |
| LMArena French | — | 1456 |
| LMArena German | — | 1446 |
| LMArena Japanese | — | 1428 |
| LMArena Korean | — | 1404 |
| LMArena Russian | — | 1448 |
| LMArena Spanish | — | 1450 |
Instruction Following Not comparable
Devstral Small 2505: —, Grok 4.5: 76.0 (#48)
| Benchmark | Devstral Small 2505 | Grok 4.5 |
|---|---|---|
| LMArena Instruction Following | — | 1446 |
Long Context Not comparable
Devstral Small 2505: —, Grok 4.5: 44.8 (#56)
| Benchmark | Devstral Small 2505 | Grok 4.5 |
|---|---|---|
| LMArena Longer Query | — | 1463 |
Writing & Preference Not comparable
Devstral Small 2505: —, Grok 4.5: 65.8 (#42)
| Benchmark | Devstral Small 2505 | Grok 4.5 |
|---|---|---|
| LMArena Text | — | 1448 |
| LMArena Creative Writing | — | 1442 |
| EQ-Bench Creative Writing | — | 1579 |
| LMArena Multi-Turn | — | 1456 |
Frequently asked questions
Is Devstral Small 2505 better than Grok 4.5?
Grok 4.5 is the stronger model overall, scoring 55.0 to 34.3 on the Noometry Index. Devstral Small 2505 costs 20× less per token, which makes it the better buy when Grok 4.5's lead doesn't matter for your workload.
Which is cheaper, Devstral Small 2505 or Grok 4.5?
Devstral Small 2505 is cheaper. It lists at $0.10 per million input tokens and $0.30 per million output tokens; Grok 4.5 lists at $2 and $6.
Is Devstral Small 2505 or Grok 4.5 better for coding?
Grok 4.5 scores higher on coding benchmarks: 52.2 versus 38.9 in the Noometry coding category.
Which has the bigger context window?
Grok 4.5 does, with 500K tokens against 128K.
How many benchmarks do Devstral Small 2505 and Grok 4.5 share?
3 benchmarks have published results for both models. Devstral Small 2505 has 4 scored results on Noometry and Grok 4.5 has 52.