Claude Opus 4.6vsClaude Opus 4
Across 11 shared benchmarks, Claude Opus 4.6 leads overall: Claude Opus 4.6 wins 10, Claude Opus 4 wins 1, with 0 ties and an average score difference of +27.08.
Claude Opus 4.6
Anthropic · 2026-02-05 · Reasoning model
Across 11 shared benchmarks, Claude Opus 4.6 leads overall: Claude Opus 4.6 wins 10, Claude Opus 4 wins 1, with 0 ties and an average score difference of +27.08.
Anthropic · 2026-02-05 · Reasoning model
Anthropic · 2025-05-23 · Reasoning model
Grouped by capability, sorted by largest gap within each. 11 shared benchmarks.
| Benchmark | Claude Opus 4.6 | Claude Opus 4 | Diff |
|---|---|---|---|
| ARC-AGI-2 | 66.3014 / 58Extended (no tools) | 8.6038 / 58 | +57.70 |
| ARC-AGI | 9211 / 65Extended (no tools) | 35.7048 / 65 | +56.30 |
| HLE | 538 / 149Extended (with tools, internet) | 10.70121 / 149 | +42.30 |
| GPQA Diamond | 91.3112 / 175Extended (no tools) | 79.6076 / 175 | +11.71 |
| Benchmark | Claude Opus 4.6 | Claude Opus 4 | Diff |
|---|---|---|---|
| FrontierMath | 40.707 / 60最高（无工具） | 4.5039 / 60 | +36.20 |
| AIME2025 | 99.797 / 106Extended (no tools) | 75.5065 / 106 | +24.29 |
| FrontierMath - Tier 4 | 22.9012 / 80最高（无工具） | 072 / 80Normal (No Tools) | +22.90 |
| MATH-500 | 97.6010 / 44Extended (no tools) |
| Benchmark | Claude Opus 4.6 | Claude Opus 4 | Diff |
|---|---|---|---|
| LiveCodeBench | 7635 / 118Extended (no tools) | 56.6074 / 118 | +19.40 |
| SWE-bench Verified | 80.846 / 103Extended (with tools) | 72.5043 / 103 | +8.34 |
| Benchmark | Claude Opus 4.6 | Claude Opus 4 | Diff |
|---|---|---|---|
| τ²-Bench | 91.891 / 40Extended (with tools) | 72.5022 / 40thinking + 使用工具 | +19.39 |
| Field | Claude Opus 4.6 | Claude Opus 4 |
|---|---|---|
| Publisher | Anthropic | Anthropic |
| Release date | 2026-02-05 | 2025-05-23 |
| Model type | Reasoning model | Reasoning model |
| Architecture | Dense | Dense |
| Parameters | 0.0 | Not available |
| Context length | 1000K | 200K |
| Max output | 65536 | 32000 |
Prices use DataLearner records when available; missing fields are not inferred.
| Item | Claude Opus 4.6 | Claude Opus 4 |
|---|---|---|
| Text input | $0.5 / 1M tokens | 15 美元/ 100万tokens |
| Text output | $25 / 1M tokens | 75 美元/100万tokens |
| Cache read | $0.5 / 1M tokens | Not public |
| Cache write | $10 / 1M tokens | Not public |
On average across the 11 shared benchmarks, Claude Opus 4.6 scores 27.08 higher.
Largest single-benchmark gap: ARC-AGI-2 — Claude Opus 4.6 66.30 vs Claude Opus 4 8.60 (+57.70).
Page generated from structured model, pricing and benchmark records. No real-time LLM is used to write the prose.
| 98.203 / 44 |
| -0.60 |