Compare / head-to-head
Claude Opus 5vsClaude Sonnet 4.6
Claude Opus 5 leads 4 of 4 shared benchmarks. Claude Sonnet 4.6 is 1.7x cheaper per token. Both offer a 1M-token context window.
Benchmarks from cited public sources; pricing from official pages; status from official provider feeds.
Shared benchmarks
4 – 0
Claude Opus 5 leads
Cheaper per token
Claude Sonnet 4.6
1.7x cheaper, input + output
Larger context
Tie
both 1M tokens
Provider uptime (30d)
100%
Anthropic
Claude Opus 5
Anthropic · released 2026-07-24
textvision
Claude Sonnet 4.6
Anthropic · released 2026-02-17
textvision
Quality
Benchmark matrix
4 shared · 5 only Claude Opus 5 · 14 only Claude Sonnet 4.6| Benchmark | Claude Opus 5 | Claude Sonnet 4.6 | Δ | Edge |
|---|---|---|---|---|
| Reported by both · 4 | ||||
| GPQA Diamond | 93.9% ↗epoch | 89.9% ↗ | +4 pt | Claude Opus 5 |
| Humanity's Last Exam (no tools) | 56.6% ↗ | 33.2% ↗ | +23.4 pt | Claude Opus 5 |
| OTIS Mock AIME 2024-2025 (Epoch AI run) | 98.9% ↗epoch | 85.8% ↗epoch | +13.1 pt | Claude Opus 5 |
| SimpleQA Verified | 59.9% ↗epoch | 35.5% ↗epoch | +24.4 pt | Claude Opus 5 |
| Only Claude Opus 5 reports · 5 | ||||
| CursorBench 3.2.0 | 70% ↗ | not reported | — | — |
| FrontierMath Tier 4 v2 (Epoch AI run) | 73.2% ↗epoch | not reported | — | — |
| FrontierMath Tiers 1-3 v2 (Epoch AI run) | 85.6% ↗epoch | not reported | — | — |
| GDPval-AA v2 | 1824 ↗ | not reported | — | — |
| Terminal-Bench Science 0.1 | 29% ↗ | not reported | — | — |
| Only Claude Sonnet 4.6 reports · 14 | ||||
| SWE-bench Verified | not reported | 79.6% ↗ | — | — |
| LMArena Elo | not reported | 1458.3 ↗ | — | — |
| ARC-AGI-2 (Verified) | not reported | 58.3% ↗ | — | — |
| GDPval-AA | not reported | 1633 ↗ | — | — |
| Humanity's Last Exam (with tools) | not reported | 49% ↗ | — | — |
| MCP Atlas | not reported | 61.3% ↗ | — | — |
| MMMLU | not reported | 89.3% ↗ | — | — |
| MMMU-Pro (no tools) | not reported | 74.5% ↗ | — | — |
| MMMU-Pro (with tools) | not reported | 75.6% ↗ | — | — |
| OSWorld-Verified | not reported | 72.5% ↗ | — | — |
| SWE-bench Verified (Epoch AI run) | not reported | 75.2% ↗epoch | — | — |
| Tau2-bench Retail | not reported | 91.7% ↗ | — | — |
| Tau2-bench Telecom | not reported | 97.9% ↗ | — | — |
| Terminal-Bench 2.0 | not reported | 59.1% ↗ | — | — |
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 5 minus Claude Sonnet 4.6 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 5 | Claude Sonnet 4.6 | Edge |
|---|---|---|---|
| Context window | 1M tokens | 1M tokens | Tie |
| Max output | 128K tokens | 128K tokens | Tie |
| Input price / 1M | $5 ↗ | $3 ↗ | Claude Sonnet 4.6 |
| Output price / 1M | $25 ↗ | $15 ↗ | Claude Sonnet 4.6 |
| Cached input / 1M | $0.5 ↗ | $0.3 ↗ | Claude Sonnet 4.6 |
| Input + output / 1M Lower is cheaper. List prices; batch, tool and regional fees excluded. | $30 | $18 | Claude Sonnet 4.6 |
| Modalities | text · vision | text · vision | Tie |
| Released | 2026-07-24 | 2026-02-17 | — |
| Cited benchmark scores | 9 | 19 | — |
Reliability
Anthropic status
Last 60 days · refreshed daily on this page · live on /statusMore matchups
Claude Opus 5 vs …
Models sharing the most benchmarksMore matchups
Claude Sonnet 4.6 vs …
Models sharing the most benchmarksBuilt by Respan
Which one wins on your data?
Public benchmarks are a starting point. Run Claude Opus 5 and Claude Sonnet 4.6 on your own prompts with Respan evals, or route to either through one gateway key with automatic failover.