Model comparison
Claude Opus 4.8 vs Devstral Small 2505
Claude Opus 4.8 is the stronger model overall, scoring 60.7 to 34.3 on the Noometry Index. Devstral Small 2505 costs 67× less per token, which makes it the better buy when Claude Opus 4.8's lead doesn't matter for your workload.
Last verified . 3 shared benchmarks.
Mistral AISummary
- They share 3 benchmarks with published results for both. Claude Opus 4.8 scores higher in 2 categories and Devstral Small 2505 in 0 categories; 2 gaps are clear of the uncertainty.
- The widest gap is in reasoning, where Claude Opus 4.8 leads 64.7 to 19.7.
- The biggest single-benchmark swing is Kagi LLM Benchmark: 88.8% for Claude Opus 4.8 and 37.7% for Devstral Small 2505.
- Devstral Small 2505 is cheaper at $0.10 / $0.30 per million input/output tokens, against $5 / $25 for Claude Opus 4.8.
- Claude Opus 4.8 accepts more context: 1M tokens versus 128K.
- Devstral Small 2505 has downloadable open weights; the other is API-only.
Side by side
| Claude Opus 4.8 | Devstral Small 2505 | |
|---|---|---|
| Provider | Anthropic | Mistral AI |
| Noometry Index | 60.7 | 34.3 |
| Released | 2026-05-28 | 2025-05-07 |
| Weights | Proprietary | Open |
| Context window | 1M | 128K |
| Max output | 128K | 128K |
| Input $ / M tokens | $5 | $0.10 |
| Output $ / M tokens | $25 | $0.30 |
| Results tracked | 65 | 4 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Claude Opus 4.8 leads
Claude Opus 4.8: 59.9 (#12), Devstral Small 2505: 38.9 (#166)
| Benchmark | Claude Opus 4.8 | Devstral Small 2505 |
|---|---|---|
| SciCode | 53.5% | 28.8% |
| DeepSWE | 59% | — |
| FrontierCode | 46.5% | — |
| SWE-bench Verified (bash only) | — | 56.4% |
| LMArena WebDev | 1556 | — |
| GSO | 47.1% | — |
| WeirdML | 82.9% | — |
| LMArena Coding | 1490 | — |
| ALE-Bench | 1,564 | — |
Agentic & Tool Use Not comparable
Claude Opus 4.8: 47.6 (#11), Devstral Small 2505: —
| Benchmark | Claude Opus 4.8 | Devstral Small 2505 |
|---|---|---|
| APEX-Agents | 48.9% | — |
| OSWorld 2.0 | 20.6% | — |
| Remote Labor Index | 8.3% | — |
| τ²-bench Banking | 39.7% | — |
| DeepResearch Bench | 50.2% | — |
| PostTrainBench | 33.8% | — |
| GBAEval | 70.9% | — |
| GDP.pdf | 24% | — |
| LMArena Search | 1204 | — |
| Vending-Bench 2 | 5,787 | — |
Reasoning Claude Opus 4.8 leads
Claude Opus 4.8: 64.7 (#16), Devstral Small 2505: 19.7 (#252)
| Benchmark | Claude Opus 4.8 | Devstral Small 2505 |
|---|---|---|
| Kagi LLM Benchmark | 88.8% | 37.7% |
| CritPt | 20.9% | 0% |
| ARC-AGI-2 | 72.1% | — |
| SimpleBench | 64.8% | — |
| NYT Connections (extended) | 91.1% | — |
| ARC-AGI-1 | 92.5% | — |
| Chess Puzzles | 34% | — |
| EnigmaEval | 23.5% | — |
| EBR-Bench | 28.6% | — |
| LMArena Hard Prompts | 1482 | — |
| Mystery Game Puzzles | 36% | — |
| DTBench | 94.9% | — |
| LMCA | 57.5% | — |
| Surface Evolver Bench | 87.5% | — |
| Bench to the Future 3 | 0.14 | — |
| Epoch Capabilities Index | 158.21 | — |
| ForecastBench | 59.9 | — |
Math Not comparable
Claude Opus 4.8: 78.4 (#13), Devstral Small 2505: —
| Benchmark | Claude Opus 4.8 | Devstral Small 2505 |
|---|---|---|
| FrontierMath (Tiers 1-3) | 80% | — |
| FrontierMath Tier 4 | 56.1% | — |
| MathArena Final-Answer Competitions | 91.8% | — |
| OTIS Mock AIME 2024-2025 | 98.3% | — |
| ProofBench | 69% | — |
| LMArena Math | 1487 | — |
| FrontierMath (Feb 2025 set) | 47.2% | — |
| FrontierMath Tier 4 (v1) | 31.3% | — |
Knowledge Not comparable
Claude Opus 4.8: 61.3 (#29), Devstral Small 2505: —
| Benchmark | Claude Opus 4.8 | Devstral Small 2505 |
|---|---|---|
| GPQA Diamond | 91% | — |
| SimpleQA Verified | 53% | — |
| LMArena Expert | 1502 | — |
Multimodal Not comparable
Claude Opus 4.8: 42.9 (#26), Devstral Small 2505: —
| Benchmark | Claude Opus 4.8 | Devstral Small 2505 |
|---|---|---|
| LMArena Vision | 1294 | — |
| Blueprint-Bench 2 | 14.5% | — |
| Furniture Assembly | 42.5% | — |
| LMArena Document | 1475 | — |
Multilingual Not comparable
Claude Opus 4.8: 55.2 (#33), Devstral Small 2505: —
| Benchmark | Claude Opus 4.8 | Devstral Small 2505 |
|---|---|---|
| LMArena Non-English | 1450 | — |
| LMArena Chinese | 1507 | — |
| LMArena French | 1481 | — |
| LMArena German | 1472 | — |
| LMArena Japanese | 1440 | — |
| LMArena Korean | 1432 | — |
| LMArena Russian | 1474 | — |
| LMArena Spanish | 1466 | — |
Instruction Following Not comparable
Claude Opus 4.8: 77.4 (#24), Devstral Small 2505: —
| Benchmark | Claude Opus 4.8 | Devstral Small 2505 |
|---|---|---|
| LMArena Instruction Following | 1476 | — |
Long Context Not comparable
Claude Opus 4.8: 45.4 (#35), Devstral Small 2505: —
| Benchmark | Claude Opus 4.8 | Devstral Small 2505 |
|---|---|---|
| LMArena Longer Query | 1483 | — |
Writing & Preference Not comparable
Claude Opus 4.8: 72.0 (#16), Devstral Small 2505: —
| Benchmark | Claude Opus 4.8 | Devstral Small 2505 |
|---|---|---|
| LMArena Text | 1461 | — |
| LMArena Creative Writing | 1454 | — |
| EQ-Bench Creative Writing | 1840 | — |
| EQ-Bench 4 | 1281 | — |
| LMArena Multi-Turn | 1476 | — |
Frequently asked questions
Is Claude Opus 4.8 better than Devstral Small 2505?
Claude Opus 4.8 is the stronger model overall, scoring 60.7 to 34.3 on the Noometry Index. Devstral Small 2505 costs 67× less per token, which makes it the better buy when Claude Opus 4.8's lead doesn't matter for your workload.
Which is cheaper, Claude Opus 4.8 or Devstral Small 2505?
Devstral Small 2505 is cheaper. It lists at $0.10 per million input tokens and $0.30 per million output tokens; Claude Opus 4.8 lists at $5 and $25.
Is Claude Opus 4.8 or Devstral Small 2505 better for coding?
Claude Opus 4.8 scores higher on coding benchmarks: 59.9 versus 38.9 in the Noometry coding category.
Which has the bigger context window?
Claude Opus 4.8 does, with 1M tokens against 128K.
How many benchmarks do Claude Opus 4.8 and Devstral Small 2505 share?
3 benchmarks have published results for both models. Claude Opus 4.8 has 65 scored results on Noometry and Devstral Small 2505 has 4.