Compare Start here

AnthropicLanguage modelSuperseded

Claude Opus 4

Launched the Claude 4 era and Claude Code's mainstream moment.

Compare head-to-head Race it in 3D

Overall Intelligence

#45

of 56
21.6 pts behind #1

0

races led

133

weeks at #1

67.1

score

Released: May 22, 2025; 1.4 yr ago
Context: 200K; 32K max output
Price: $15 / $75; per 1M in/out · $30.00 blended
Speed: 42 tok/s; 4s to first token
Modalities: text · image; outputs text
Weights: Closed; API and apps

The 30-second version

Launched the Claude 4 era and Claude Code's mainstream moment.

Key takeaways

1Launched the Claude 4 era and Claude Code's mainstream moment.

Where it ranks

Current snapshot · illustrative scores · click any race to watch it

Opus 4Best on each axis

Overall

45Overall+21.6 pts

Coding

42Coding+20.9 pts 38Frontend & UI+116 Elo 42Backend & APIs+18.4 pts 44Distributed Systems+23.3 pts 41Full-Stack Apps+19.3 pts 42Debugging & Repair+21.6 pts 41Agentic Coding+22.1 pts 42Code Review & Security+18.4 pts 42Data Science & SQL+18.3 pts 41Mobile Apps+18.3 pts 40Terminal & DevOps+24.2 pts 45Competitive Programming+676

Hard Reasoning

44Hard Reasoning+22.7 pts 46Frontier Questions+42.3 pts 42Abstract & Novel Reasoning+32.8 pts 46Logic & Puzzles+20.8 pts 40Planning & Strategy+18.3 pts

Math

48Math+24.9 pts 51Competition Math+18.9 pts 48Research Math+32.2 pts 46Applied & Quantitative+20.8 pts

Science

45Science+22.9 pts 43Graduate Science QA+12.9 pts 44Biology & Medicine+19.7 pts 46Physics & Chemistry+21.3 pts 45Research Assistance+21.1 pts

Writing & Language

42Writing & Language+20.3 pts 41Creative Writing+122 Elo 41Technical Writing+15.0 pts 42Editing & Style+16.2 pts 44Multilingual+14.3 pts 43Instruction Following+15.3 pts 43Conversation & EQ+124 Elo

Multimodal

41Multimodal+35.7 pts 39Image Understanding+21.8 pts 38Documents & Charts+20.2 pts 41Video Understanding+37.6 pts 41Speech & Audio+45.6 pts

Tool Use & Agents

42Tool Use & Agents+22.4 pts 42Tool Use+15.9 pts 38Computer Use+29.7 pts 44Web Research+28.2 pts 44Long-Horizon+27.0 hrs

Long Context

45Long Context+26.4 pts 471M Recall+22.4 pts 44Long Reasoning+22.9 pts

Speed & Cost

55Speed & Cost+62.5 pts 53Output Speed+480 tok/s 28Latency+3.69s 56Cost-Efficiency+83.3 pts

Trust & Factuality

42Trust & Factuality+21.3 pts 42Factuality+16.7 pts 43Honesty & Calibration+18.3 pts

Overall Intelligence over time

#1 frontierOpus 4V4.1 FlashGLM-5.3

Update history

The distilled version first; open any release for the deeper breakdown.

The Claude Opus family

Opus 4May 2025 Opus 4.1August 2025 Opus 4.5November 2025 Opus 4.6February 2026 Opus 4.8May 2026 Opus 5July 2026 Opus 5.5September 2026

60-second briefing

Opus 4, explained

AI video summary · placeholder render from the auto-written script

From Anthropic's own guide, distilled

Prompt engineering and Claude best practices

Be explicit about the outcome you want; newer Claude models follow instructions literally rather than guessing at hidden intent.
Explain why a rule exists. Context about the goal lets the model generalise the rule to cases you did not list.
Wrap documents, examples and data in XML-style tags so instructions and material never blur together.
Put long documents at the top of the prompt and your question at the end; recall and grounding both improve.
For agentic tasks, state when to act versus when to ask, and ask for a short plan before large edits.

All official guides, distilled →

About Anthropic

Safety-focused lab building the Claude family of models.

Optimising for

Reliable long-horizon agentic work (especially coding and computer use), honest and calibrated answers, and strong writing, with safety research shipped alongside capability.

San Francisco, US · founded 2021

Closest rivals

Opus 4 is 45th in overall intelligence; leader is Claude Opus 5.5.