Model dossier模型档案
Claude Opus 5
Claude Opus 5 is a high-tier model from Anthropic (Intelligence Score 83.3/100). Strongest in reasoning (96.4) and knowledge (88.5); best fit: research, analysis and planning. Premium-priced: $5.00 in / $25 out per 1M tokens. Context window 1M tokens — long-document friendly.
Claude Opus 5 是 Anthropic 的高水平梯队模型(智能评分 83.3/100)。 最强项是推理(96.4分),其次是知识(88.5分);适合研究、分析与规划。 定价属高端定价:每百万 token 输入 $5.00 / 输出 $25。 上下文 1M token,长文档友好。
83.3
Intelligence Score · benchmark coverage 智能评分 · benchmark 覆盖 100%
◈
Benchmark scoresBenchmark 成绩
11 entries 条| Benchmark基准 | Score成绩 | Index指数 | Dated日期 | Measured by测评方 | |
|---|---|---|---|---|---|
| Humanity's Last ExamReasoning | 54.4⚡max | 95 | — | 3rd-party第三方 | source ↗×5⚠±12.2 |
| GPQA DiamondReasoning | 93.7 | 98 | — | 3rd-party第三方 | source ↗×11⚠±5.2 |
| ↳ ⚡low | 93 | — | — | 3rd-party第三方 | source ↗ |
| MMLU-ProKnowledge | 81.97 | 89 | — | 3rd-party第三方 | source ↗×4⚠±10.53 |
| SWE-bench VerifiedCoding | 96 | 100 | — | official官方榜 | source ↗×10 |
| AIME 2025Math | 97 | 97 | — | 3rd-party第三方 | source ↗×4⚠±27.1 |
| LiveCodeBenchCoding | 89.03 | 95 | 2026-07 | 3rd-party第三方 | source ↗×5⚠±29.63 |
| FrontierMathMath | 31.25 | 33 | — | 3rd-party第三方 | source ↗×5⚠±56.75 |
| Terminal-BenchCoding | 43.5⚡max | 47 | — | official官方榜 | source ↗×7⚠±46.4 |
| τ²-benchAgent | 48.71⚡max | 49 | 2026-08-04 | official官方榜 | source ↗×3⚠±34.07 |
| LMArena (Chatbot Arena)Preference | 1489⚡max | 67 | 2026-08-12 | official官方榜 | source ↗×10⚠±210 |
| ↳ ⚡high | 1493 | — | 2026-08-12 | official官方榜 | source ↗ |
| WebArenaAgent | 68.7 | 75 | — | 3rd-party第三方 | source ↗×2 |
≡
Specs & pricing规格与价格
| Provider厂商 | Anthropic |
|---|---|
| Released发布 | 2026-07-24 |
| Context window上下文 | 1M tokens |
| Max output最大输出 | 128K tokens |
| Modality模态 | text+image+file->text |
| Reasoning tiers推理档位 | ⚡low ⚡medium ⚡high ⚡max default默认 high · low / medium / high / x-high |
| API input price输入价格 | $5.00 / 1M tokens |
| API output price输出价格 | $25 / 1M tokens |
⛓
Tools using this model使用它的工具
⚔
Head-to-head正面对比
How we score →评分方法 → · Catalog & pricing via the official OpenRouter API.模型目录与价格来自 OpenRouter 官方 API。