DataLearner logo
OP

Opus 4.7

Reasoning modelOpusClaude 4.7

Claude Opus 4.7

Release date: 2026-04-16Updated: 2026-06-15 07:18:23.302Knowledge cutoff: 2026-015,693
Live demoGitHubHugging FaceCompare
Parameters
Not disclosed
Context length
1000K
Chinese support
Supported
Reasoning ability

Claude Opus 4.7 是 Anthropic 于 2026 年 4 月 16 日发布的旗舰推理大模型，是 Opus 4 系列的最新迭代版本，支持长达 100 万 token（1000K）的超长上下文窗口，最大输出长度为 65,536 tokens。模型不开源，可通过 claude.ai 在线体验或调用 Anthropic API 使用。

Data sourced primarily from official releases (GitHub, Hugging Face, papers), then benchmark leaderboards, then third-party evaluators. Learn about our data methodology

Opus 4.7

Model basics

Reasoning traces
Supported
Thinking modes
Thinking Level · Extended (Default)Standard ModeThinking Level · MediumThinking Level · HighThinking Level · Extra-High
Context length
1000K tokens
Max output length
128K tokens
Model type
Reasoning model
Modality (in / out)
Text, Image → Text
Release date
2026-04-16
Model file size
No data
MoE architecture
No
Total params / Active params
No data / N/A
Knowledge cutoff
2026-01
Opus 4.7

Open source & experience

Code license
不开源
Weights license
不开源
GitHub repo
GitHub link unavailable
Hugging Face
Hugging Face link unavailable
Live demo
https://claude.ai/
Opus 4.7

Official resources

Opus 4.7

API details

API speed
3/5
💡Default unit: $/1M tokens. If vendors use other units, follow their published pricing.
Learn about pricing modes
Standard
TypeConditionInputOutput
Text-$5.00/ 1M$25.00/ 1M
Batch
TypeConditionInputOutput
Text-$2.50/ 1M$12.50/ 1M
Cache PricingPrompt Cache
TypeTTLWriteRead
Text5m$6.25/ 1M$0.500/ 1M
Text1h$10.00/ 1M$0.500/ 1M
Opus 4.7

Benchmark Results

Opus 4.7 currently shows benchmark results led by GPQA Diamond (4 / 180, score 94.20), SWE-bench Verified (5 / 109, score 87.60), LiveBench (7 / 115, score 76.91). This page also consolidates core specs, context limits, and API pricing so you can evaluate the model from benchmark results and deployment constraints together.

Thinking
Tool usage

General Knowledge

17 evaluations
Benchmark / mode
Score
Rank/total
GPQA Diamond
Extended
94.20
4 / 180
ARC-AGI
Low
91
14 / 67
ARC-AGI
Medium
91
14 / 67
ARC-AGI
High
93.50
10 / 67
ARC-AGI
Max
92
12 / 67
MMLU
Standard Mode
91.50
6 / 65
LiveBench
Low
70.09
39 / 115
LiveBench
Medium
72.31
27 / 115
LiveBench
High
74.89
18 / 115
LiveBench
Deep
76.91
7 / 115
ARC-AGI-2
Low
62.10
18 / 61
ARC-AGI-2
Medium
67.50
15 / 61
ARC-AGI-2
High
68.30
14 / 61
ARC-AGI-2
Max
75.80
10 / 61
HLE
Extended
46.90
34 / 164
HLE
ExtendedTools
54.70
10 / 164
ARC-AGI-3
High
0
7 / 8

Coding and Software Engineer

2 evaluations
Benchmark / mode
Score
Rank/total
SWE-bench Verified
ExtendedTools
87.60
5 / 109
SWE-Bench Pro - Public
ExtendedTools
64.30
6 / 48

常识推理

1 evaluations
Benchmark / mode
Score
Rank/total
Simple Bench
Standard Mode
61.70
13 / 63

Math and Reasoning

2 evaluations
Benchmark / mode
Score
Rank/total
FrontierMath
Extra-High
43.80
6 / 60
FrontierMath - Tier 4
Extra-High
22.90
12 / 80

AI Agent - Information Search

1 evaluations
Benchmark / mode
Score
Rank/total
BrowseComp
ExtendedTools
79.30
15 / 49

AI Agent - Tool Usage

4 evaluations
Benchmark / mode
Score
Rank/total
MCP-Atlas
DeepTools
79.10
6 / 25
OSWorld-Verified
ExtendedTools
78
8 / 20
TerminalBench 2.1
HighTools
69.70
16 / 21
Terminal Bench 2.0
ExtendedTools
69.40
6 / 46
View benchmark analysisCompare with other models

Compare with other models

Opus 4.7

Publisher

Claude Opus 4.7

Model Overview

Claude Opus 4.7 是 Anthropic 于 2026 年 4 月 16 日发布的旗舰推理大模型，是 Opus 4 系列的最新迭代版本，支持长达 100 万 token（1000K）的超长上下文窗口，最大输出长度为 65,536 tokens。模型不开源，可通过 claude.ai 在线体验或调用 Anthropic API 使用。

DataLearner on WeChat

Follow DataLearner on WeChat for AI model updates and research notes.

DataLearner WeChat QR code