See key specs and per-benchmark scores for each model/mode. Scroll horizontally for all columns. 当前对比 2 个模型的评测数据与核心参数。
Claude Sonnet 4.5
Anthropic
Each axis is a category average, normalized to a 100-point radar.
Relative edge: none clear / Relative gap: AI Agent - 信息收集 -50.6
Relative edge: AI Agent - 信息收集 +50.6 / Relative gap: none clear
Method: for each model and benchmark, the chart first averages all scores in the current mode scope instead of taking the best score, then averages those benchmark scores within each category. Only benchmarks with at least two selected models scored are included; missing values are not counted as zero.
Best overall
Claude Sonnet 4.6 · 67.88
Best single
Claude Sonnet 4.5 · τ²-Bench - Telecom 98.00
Modality coverage
Claude Sonnet 4.5 · 2 modalities
Head to head
14
Benchmarks
2
Wins
12
Losses
-13.65
Average diff
Compare benchmark results across thinking modes and tool usage.
Data sourced primarily from official releases (GitHub, Hugging Face, papers), then benchmark leaderboards, then third-party evaluators. Learn about our data methodology
Complete scores for each model/mode across selected benchmarks.
14 benchmarks with comparable scores. Each model shows its best score; mode label is displayed below.
| Benchmark | Claude Sonnet 4.5 | Claude Sonnet 4.6 |
|---|---|---|
ARC-AGI-2 综合评估 | 13.60Thinking Enabled | 58.30Thinking Enabled |
GPQA Diamond 综合评估 | 83.40Thinking Enabled | 89.90Thinking Enabled |
HLE 综合评估 | 33.60Thinking Enabled ｜ Tools | 49.00Thinking Enabled ｜ Tools |
LiveBench 综合评估 | 68.1964K | 75.47Thinking Level · Medium |
SWE-bench Verified 编程与软件工程 | 77.20Thinking Enabled ｜ Tools | 79.60Thinking Enabled |
4.2032K | 8.3016K | |
MCP-Atlas AI Agent - 工具使用 | 59.50Thinking Enabled ｜ Tools | 69.50Standard Mode ｜ Tools |
OSWorld-Verified AI Agent - 工具使用 | 61.40Thinking Enabled ｜ Tools | 72.50Thinking Enabled ｜ Tools |
Terminal Bench 2.0 AI Agent - 工具使用 | 42.80Thinking Enabled ｜ Tools | 59.10Thinking Enabled ｜ Tools |
τ²-Bench - Telecom Agent能力评测 | 98.00Thinking Enabled ｜ Tools | 97.90Thinking Enabled ｜ Tools |
BrowseComp AI Agent - 信息收集 | 24.10Thinking Enabled ｜ Tools | 74.70Thinking Enabled ｜ Tools |
GDPval-AA 生产力知识 | 39.00Thinking Enabled | 57.00Thinking Enabled |
AA-LCR 长上下文能力 | 66.00Thinking Enabled | 71.00Thinking Enabled |
Pinch Bench OpenClaw智能体能力综合测评 | 88.20Thinking Enabled ｜ Tools | 88.00Thinking Enabled ｜ Tools |
Side-by-side input/output token pricing
Licensing, MoE architecture, and multi-modality support.
| Features & specs | Claude Sonnet 4.5Anthropic | Claude Sonnet 4.6Anthropic |
|---|---|---|
Core specsRelease | 2025-09-30 | 2026-02-17 |
Context length | 1000K | 1M |
Max output | 65536 | 8192 |
MoE | No | No |
Supported modes | 常规模式（Non-Thinking Mode）思考模式（Thinking Mode）深度思考（Deeper Thinking Mode） | 常规模式（Non-Thinking Mode）思考模式（Thinking Mode） |
LicenseCode Open Source | Not provided | Not provided |
Weights Open Source | Not provided | Not provided |
Commercial use | 不开源 | 不开源 |
Modality supportText Input/Output | / | / |
Image Input/Output | / | / |
ResourcesPaper / report | Introducing Claude Sonnet 4.5 | Introducing Claude Sonnet 4.6 |
DataLearner blog | 全球最强编程大模型升级：Anthropic发布Claude Sonnet 4.5！同时还有一波重磅工具更新：Claude Code支持保存状态等 | Not provided |
Claude Sonnet 4.6
Anthropic
