Compare / head-to-head
Claude Opus 4.5vsClaude Sonnet 4.5
Claude Opus 4.5 leads 23 of 25 shared benchmarks. Claude Sonnet 4.5 is 1.7x cheaper per token. Both offer a 200K-token context window.
Benchmarks from cited public sources; pricing from official pages; status from official provider feeds.
Shared benchmarks
23 – 1
Claude Opus 4.5 leads · 1 tied
Cheaper per token
Claude Sonnet 4.5
1.7x cheaper, input + output
Larger context
Tie
both 200K tokens
Providers
Anthropic
same provider
Claude Opus 4.5
Anthropic · released 2025-11-24
textvision
Claude Sonnet 4.5
Anthropic · released 2025-09-29
textvision
Quality
Benchmark matrix
25 shared · 13 only Claude Opus 4.5 · 11 only Claude Sonnet 4.5| Benchmark | Claude Opus 4.5 | Claude Sonnet 4.5 | Δ | Edge |
|---|---|---|---|---|
| Reported by both · 25 | ||||
| SWE-bench Verified | 80.9% ↗ | 77.2% ↗ | +3.7 pt | Claude Opus 4.5 |
| GPQA Diamond | 87% ↗ | 83.4% ↗ | +3.6 pt | Claude Opus 4.5 |
| LMArena Elo | 1473.6 ↗ | 1456.6 ↗ | +17 | Claude Opus 4.5 |
| ARC-AGI-2 (Verified) | 37.6% ↗ | 13.6% ↗ | +24 pt | Claude Opus 4.5 |
| Chess Puzzles (Epoch AI run) | 12% ↗epoch run | 12% ↗epoch run | 0 | Tie |
| EBR-bench (Epoch AI run) | 14.3% ↗epoch run | 2.4% ↗epoch run | +11.9 pt | Claude Opus 4.5 |
| FrontierMath Tier 4 v2 (Epoch AI run) | 4.9% ↗epoch run | 2.4% ↗epoch run | +2.5 pt | Claude Opus 4.5 |
| FrontierMath Tiers 1-3 v2 (Epoch AI run) | 34.4% ↗epoch run | 23.9% ↗epoch run | +10.5 pt | Claude Opus 4.5 |
| GSO Opt@1 (GSO) | 24.5% ↗ | 12.7% ↗ | +11.8 pt | Claude Opus 4.5 |
| LMArena WebDev (LMArena) | 1493.3 ↗ | 1393 ↗ | +100.3 | Claude Opus 4.5 |
| MCP Atlas | 62.3% ↗ | 43.8% ↗ | +18.5 pt | Claude Opus 4.5 |
| MMMLU | 90.8% ↗ | 89.5% ↗ | +1.3 pt | Claude Opus 4.5 |
| Mystery Game Puzzles (Epoch AI run) | 22% ↗epoch run | 17% ↗epoch run | +5 pt | Claude Opus 4.5 |
| OTIS Mock AIME 2024-2025 (Epoch AI run) | 86.1% ↗epoch run | 77.8% ↗epoch run | +8.3 pt | Claude Opus 4.5 |
| SAGE (Vals AI) | 52.1% ↗ | 36.1% ↗ | +16 pt | Claude Opus 4.5 |
| SimpleBench (SimpleBench) | 62% ↗ | 54.3% ↗ | +7.7 pt | Claude Opus 4.5 |
| SimpleQA Verified | 45.7% ↗epoch run | 30.7% ↗epoch run | +15 pt | Claude Opus 4.5 |
| SWE-bench Verified (Epoch AI run) | 76.7% ↗epoch run | 71.3% ↗epoch run | +5.4 pt | Claude Opus 4.5 |
| tau2-bench Airline (Sierra) | 84% ↗ | 72% ↗ | +12 pt | Claude Opus 4.5 |
| tau2-bench Banking Knowledge (Sierra) | 24.7% ↗ | 25.3% ↗ | -0.6 pt | Claude Sonnet 4.5 |
| tau2-bench Retail (Sierra) | 79.6% ↗ | 72.4% ↗ | +7.2 pt | Claude Opus 4.5 |
| tau2-bench Telecom (Sierra) | 92.3% ↗ | 84.9% ↗ | +7.4 pt | Claude Opus 4.5 |
| Terminal-Bench 2.0 | 59.3% ↗ | 51% ↗ | +8.3 pt | Claude Opus 4.5 |
| Vending-Bench 2 (Andon Labs) | 4967.06 ↗ | 3838.74 ↗ | +1128.3 | Claude Opus 4.5 |
| WeirdML (Håvard Tveit Ihle) | 63.7% ↗ | 47.7% ↗ | +16 pt | Claude Opus 4.5 |
| Only Claude Opus 4.5 reports · 13 | ||||
| BALROG (BALROG) | 43.5% ↗ | not reported | — | — |
| Furniture Assembly (Epoch AI run) | 28.3% ↗epoch run | not reported | — | — |
| LiveBench Agentic Coding (LiveBench) | 39.7% ↗ | not reported | — | — |
| LiveBench Coding (LiveBench) | 79.7% ↗ | not reported | — | — |
| LiveBench Data Analysis (LiveBench) | 74.4% ↗ | not reported | — | — |
| LiveBench Instruction Following (LiveBench) | 62.5% ↗ | not reported | — | — |
| LiveBench Language (LiveBench) | 81.3% ↗ | not reported | — | — |
| LiveBench Mathematics (LiveBench) | 90.4% ↗ | not reported | — | — |
| LiveBench Reasoning (LiveBench) | 80.1% ↗ | not reported | — | — |
| MMMU (validation) | 80.7% ↗ | not reported | — | — |
| OSWorld | 66.3% ↗ | not reported | — | — |
| τ2-bench (Retail) | 88.9% ↗ | not reported | — | — |
| τ2-bench (Telecom) | 98.2% ↗ | not reported | — | — |
| Only Claude Sonnet 4.5 reports · 11 | ||||
| AIME 2025 | not reported | 84.2% ↗matharena ⚠ | — | — |
| Humanity's Last Exam (no tools) | not reported | 17.7% ↗ | — | — |
| GDPval-AA | not reported | 1276 ↗ | — | — |
| Humanity's Last Exam (with tools) | not reported | 33.6% ↗ | — | — |
| MATH Level 5 (Epoch AI run) | not reported | 97.7% ↗epoch run | — | — |
| MMMU-Pro (no tools) | not reported | 63.4% ↗ | — | — |
| MMMU-Pro (with tools) | not reported | 68.9% ↗ | — | — |
| OSWorld-Verified | not reported | 61.4% ↗ | — | — |
| OSWorld-Verified (XLANG) | not reported | 62.9% ↗ | — | — |
| Tau2-bench Retail | not reported | 86.2% ↗ | — | — |
| Tau2-bench Telecom | not reported | 98% ↗ | — | — |
Scores tagged "epoch" or "matharena" are independent runs, used only where the lab has not published its own; ⚠ marks rows MathArena flags as released after the competition. Higher is better on every row. Δ is Claude Opus 4.5 minus Claude Sonnet 4.5 in the benchmark's own unit. "Not reported" means the lab has not published that figure; it is not a zero. ↗ opens the source.
Specs & pricing
Side by side
Official pricing pages and model cards| Spec | Claude Opus 4.5 | Claude Sonnet 4.5 | Edge |
|---|---|---|---|
| Context window | 200K tokens | 200K tokens | Tie |
| Max output | 64K tokens | 64K tokens | Tie |
| Input price / 1M | $5 ↗ | $3 ↗ | Claude Sonnet 4.5 |
| Output price / 1M | $25 ↗ | $15 ↗ | Claude Sonnet 4.5 |
| Cached input / 1M | $0.5 ↗ | $0.3 ↗ | Claude Sonnet 4.5 |
| Input + output / 1M Lower is cheaper. List prices; batch, tool and regional fees excluded. | $30 | $18 | Claude Sonnet 4.5 |
| Modalities | text · vision | text · vision | Tie |
| Released | 2025-11-24 | 2025-09-29 | — |
| Cited benchmark scores | 39 | 43 | — |
Reliability
Anthropic status
Live from /statusMore matchups
Claude Opus 4.5 vs …
Models sharing the most benchmarksMore matchups
Claude Sonnet 4.5 vs …
Models sharing the most benchmarksBuilt by Respan
Which one wins on your data?
Public benchmarks are a starting point. Run Claude Opus 4.5 and Claude Sonnet 4.5 on your own prompts with Respan evals, or route to either through one gateway key with automatic failover.