Comparing 2 models
Claude Sonnet 5.5 vs Qwen3.8 27B
Benchmarks through
Verdict
Claude Sonnet 5.5 and Qwen3.8 27B are not yet ranked on the same task, so compare the specifications below. Qwen3.8 27B costs 50.0x less per input token. Only Qwen3.8 27B has open weights for self-hosting.
- Choose Qwen3.8 27B for you need to self-host.
Built from the published task rankings below; ranks come from each task's own weighted benchmarks, so a win on one task says nothing about another.
Measured head to head
Benchmarks both models were run on
| Evaluation | Claude Sonnet 5.5 | Qwen3.8 27B | Source |
|---|---|---|---|
| LiveBench Agentic Coding | 56.3% | 61.4% | source · 2026-10-18 |
| LiveBench Coding | 91.4% | 75.7% | source · 2026-10-18 |
| LiveBench Data Analysis | 59.5% | 76.6% | source · 2026-10-18 |
| LiveBench Instruction Following | 56.8% | 72.7% | source · 2026-10-18 |
| LiveBench Language | 78.0% | 74.3% | source · 2026-10-18 |
| LiveBench Math | 96.1% | 86.2% | source · 2026-10-18 |
| LiveBench Reasoning | 91.6% | 80.0% | source · 2026-10-18 |
| LMArena Agent | 12.7 | -3.8 | source · 2026-10-20 |
| LMArena Agent Steerability | 10.9 | -3.9 | source · 2026-10-20 |
| LMArena Agent Task Outcome | 20.6 | -1.5 | source · 2026-10-20 |
| LMArena Agent Tool Hallucination | 0.6 | -0.1 | source · 2026-10-20 |
| LMArena Creative Writing | 1467 Elo | 1383 Elo | source · 2026-10-20 |
| LMArena Instruction Following | 1492 Elo | 1436 Elo | source · 2026-10-20 |
| LMArena Long Query | 1495 Elo | 1448 Elo | source · 2026-10-20 |
| LMArena Math | 1497 Elo | 1455 Elo | source · 2026-10-20 |
| LMArena Text | 1477 Elo | 1437 Elo | source · 2026-10-20 |
| LMArena Text Factuality | 1477 Elo | 1437 Elo | source · 2026-10-20 |
| LMArena Vision | 1265 Elo | 1242 Elo | source · 2026-10-19 |
| LMArena Vision Diagrams | 1321 Elo | 1270 Elo | source · 2026-10-19 |
| LMArena Vision OCR | 1282 Elo | 1262 Elo | source · 2026-10-19 |
| LMArena WebDev | 1768 Elo | 1593 Elo | source · 2026-10-19 |
| LMArena WebDev Image-to-Web | 1737 Elo | 1577 Elo | source · 2026-10-19 |
Values are as published by each evaluator, in its own units; rows from different evaluators are not on a common scale.
task evidence
How they place when the question changes
Lower ranks are better. “Best” marks the strongest option in this shortlist for that task.