Up to four models
Qwen3.8-27B vs Claude Opus 4.5
On the index (GPT-4 = 100), Qwen3.8-27B scores 155 and Claude Opus 4.5 152. Side by side on the same third-party tests, with price and context. The link in your address bar shares this exact comparison.
155139–171
WeightsOpen
ReleasedAug 2026
Price in / out$0.060 / $2.2
Context1.02M
Core tests taken3 of 9
Claude Opus 4.5
152136–168
WeightsClosed
ReleasedNov 2025
Price in / out$5.0 / $25
Context200K
Core tests taken5 of 9
Qwen3.8-27B and Claude Opus 4.5 are too close to call (3 points apart, within the uncertainty), and costs 17× less.
Test by testthird-party results
Every result as a table
| Test | Qwen3.8-27B | Claude Opus 4.5 |
|---|---|---|
| Reasoning & knowledge | ||
| LiveBench Data Analysis | 76.6% | 74.4% |
| LiveBench IF | 72.7% | 62.5% |
| LiveBench Language | 74.4% | 81.3% |
| LiveBench Reasoning | 80.0% | 80.1% |
| GPQA Diamond (Vals AI) | 88.9% | 85.9% |
| MMLU Pro | 84.3% | 87.3% |
| MMMU Pro | 83.9% | 83.0% |
| GeoBench | 38.0% | 75.0% |
| SimpleBench | 60.2% | 62.0% |
| Maths | ||
| LiveBench Mathematics | 86.2% | 90.4% |
| Coding | ||
| LiveBench Agentic Coding | 61.4% | 39.7% |
| LiveBench Coding | 75.7% | 79.7% |
| LiveCodeBench | 84.0% | 83.7% |
| SWE-bench | 86.0% | 76.4% |
| Vibe Code Bench v1.1 | 64.8% | 20.6% |
| Agents | ||
| DTBench | 88.0% | 89.9% |
| LMCA | 41.4% | 44.5% |
| Professional work | ||
| LegalBench | 82.4% | 84.6% |
| MedCode | 28.7% | 49.2% |
| MedScribe | 83.9% | 85.3% |
| MortgageTax | 64.9% | 68.7% |
| TaxEval v2 | 70.9% | 74.9% |
Task length · METR45 sec
Tests at least two of these models have taken. Bold rows are the core tests.