Claude Opus 4.6
Anthropic·5 个数据来源·更新于 2026-11-10
评测表现
只在同一来源、同一口径下比较。切换来源可查看不同测试结果。
不同来源的分数不能直接相加或横向比较;评测任务成本不是 API Token 单价。
逐项基准成绩
27 项已收录成绩 · 保留原始单位和测试条件
Consensus@64 accuracy (most commonly occurring answer across 64 samples). Run independently by Artificial Analysis.
成绩出处Web search, web fetch, programmatic tool calling, context compaction triggered at 50k tokens up to 10M total tokens, max reasoning effort, no thinking enabled. Multi-agent harness increased score to 86.8%.
成绩出处CharXiv Reasoning. With adaptive thinking, image cropping tool, and max effort. Scores averaged over 5 runs. Without image cropping tool: 68.7%.
成绩出处Cybersecurity vulnerability reproduction. No thinking, default effort, temperature, and top_p. Given a 'think' tool for interleaved thinking in multi-turn evaluations.
成绩出处F1 score. 900-prompt benchmark for evaluating agents on difficult multi-step information-seeking tasks. Programmatic search tools, context compaction, adaptive thinking, max effort up to 10M total tokens.
成绩出处LAB-Bench FigQA. With adaptive thinking, image cropping tool, and max effort. Scores averaged over 5 runs. Without image cropping tool (reasoning only): 58.0%.
成绩出处GraphWalks BFS 256K subset of 1M. F1 score with 64k output tokens, 1M context window, average of 5 trials. Max output tokens: 61.1. Full 1M variant: 41.2 (64k output).
成绩出处GraphWalks Parents 256K subset of 1M. F1 score with max output tokens, 1M context window, average of 5 trials. At 64k output tokens: 95.1. Full 1M variant: 72.0 (max), 71.1 (64k).
成绩出处With tools: web search, web fetch, code execution, programmatic tool calling, context compaction triggered at 50k tokens up to 3M total tokens, max reasoning effort, adaptive thinking enabled. Domain blocklist used for decontamination. Without tools: 40.0%.
成绩出处Scaled tool use. At high effort, reached industry-leading score of 62.7%. At max effort, score is higher.
成绩出处MRCR v2 1M 8-needles variant. Mean Match Ratio with max output tokens, average of 5 trials. Anthropic also reports 78.3 with 64k output tokens and 93.0 for the 256K 8-needle subset.
成绩出处Software failure diagnosis. 1 point if all generated root-cause elements match ground-truth, 0 points if any mismatch. Run on benchmark author's harness, graded using official methodology.
成绩出处Averaged over 25 trials, each run with adaptive thinking, max effort, default sampling settings (temperature, top_p), and with thinking blocks included in sampling results.
成绩出处Averaged over 25 trials. With a prompt modification, score reached 81.42%.
成绩出处Terminus-2 harness. 1× guaranteed / 3× ceiling resource allocation, 5–15 samples per task across staggered batches.
成绩出处Final balance in USD. Simulated vending machine business over a full year of operation. Starting balance $5,000.
成绩出处技术规格
只展示有可靠来源的资料;优先采用厂商官网与官方模型仓库。
价格与用量
美元 / 百万 Token。不同服务商或测试配置的报价分别列出。
| 来源 / 服务商 | 输入 | 输出 | 缓存读取 |
|---|---|---|---|
| 厂商官网 · Claude Opus 4.6 | $5 | $25 | $0.5 |
| Anthropic | $5 | $25 | 待补充 |
每个数字,都有出处
同名模型通过开发机构和明确版本关联。预览版、日期版本和不同规模模型分别建档。
厂商官网
厂商官方文档 · 基础规格与标准 API 报价
LLM Stats
公开模型页与公开榜单;自报成绩单独标识
BenchLM
公开评测数据;保留来源原始口径
Arena
Arena 官方公开数据集 · CC BY 4.0
OpenRouter
公开模型目录与 API 报价
查看全部配置与采集记录
| 来源 / 配置 | 采集时间 | 原始数据 |
|---|---|---|
| LLM StatsClaude Opus 4.6SHA-256 6bddf449beabf1c1… | 2026-11-10T13:53:59Z | 查看来源 |
| BenchLMClaude Opus 4.6 | 2026-11-10 | 查看来源 |
| ArenaClaude Opus 4.6 (High) | 2026-11-10T14:07:52Z | 查看来源 |
| ArenaClaude Opus 4.6 | 2026-11-10T14:07:52Z | 查看来源 |
| ArenaClaude Opus 4.6 (High) | 2026-11-10T14:07:52Z | 查看来源 |
| ArenaClaude Opus 4.6 (High) | 2026-11-10T14:07:52Z | 查看来源 |
| ArenaClaude Opus 4.6 | 2026-11-10T14:07:52Z | 查看来源 |
| ArenaClaude Opus 4.6 | 2026-11-10T14:07:52Z | 查看来源 |
| ArenaClaude Opus 4.6 (High) | 2026-11-10T14:07:52Z | 查看来源 |
| ArenaClaude Opus 4.6 | 2026-11-10T14:07:52Z | 查看来源 |
| OpenRouterClaude Opus 4.6SHA-256 63cbc45064d84060… | 2026-11-10T13:51:49Z | 查看来源 |
| 厂商官网Claude Opus 4.6SHA-256 b4523e3290954239… | 2026-11-10T14:06:49Z | 查看来源 |