Compare Start here

AnthropicLanguage modelSuperseded

Claude 3.7 Sonnet

The first hybrid reasoning Claude and the model that made AI coding mainstream.

Compare head-to-head Race it in 3D

Overall Intelligence

#52

of 56
27.3 pts behind #1

0

races led

74

weeks at #1

61.4

score

Released: Feb 24, 2025; 1.6 yr ago
Context: 200K; 64K max output
Price: $3 / $15; per 1M in/out · $6.00 blended
Speed: 75 tok/s; 1.6s to first token
Modalities: text · image; outputs text
Weights: Closed; API and apps

The 30-second version

The first hybrid reasoning Claude: one model that could answer instantly or think step by step.

Key takeaways

1Introduced extended thinking to Claude.
2Launched alongside the first version of Claude Code.

Where it ranks

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

3.7 SonnetBest on each axis

Overall

52Overall+27.3 pts

Coding

49Coding+27.3 pts 46Frontend & UI+149 Elo 50Backend & APIs+23.8 pts 51Distributed Systems+30.1 pts 50Full-Stack Apps+25.9 pts 49Debugging & Repair+27.7 pts 49Agentic Coding+27.5 pts 51Code Review & Security+25.6 pts 48Data Science & SQL+24.9 pts 48Mobile Apps+21.9 pts 50Terminal & DevOps+31.5 pts 54Competitive Programming+908

Hard Reasoning

54Hard Reasoning+29.6 pts 54Frontier Questions+50.5 pts 49Abstract & Novel Reasoning+43.0 pts 54Logic & Puzzles+27.4 pts 50Planning & Strategy+25.7 pts

Math

54Math+32.5 pts 54Competition Math+25.2 pts 54Research Math+43.0 pts 54Applied & Quantitative+25.8 pts

Science

53Science+28.8 pts 53Graduate Science QA+17.3 pts 52Biology & Medicine+24.4 pts 54Physics & Chemistry+26.0 pts 52Research Assistance+26.2 pts

Writing & Language

50Writing & Language+26.7 pts 50Creative Writing+154 Elo 50Technical Writing+21.3 pts 49Editing & Style+20.7 pts 51Multilingual+18.2 pts 52Instruction Following+20.1 pts 51Conversation & EQ+155 Elo

Multimodal

43Multimodal+43.4 pts 41Image Understanding+25.8 pts 41Documents & Charts+25.4 pts 43Video Understanding+46.2 pts 46Speech & Audio+54.9 pts

Tool Use & Agents

49Tool Use & Agents+28.6 pts 50Tool Use+21.7 pts 42Computer Use+37.0 pts 52Web Research+35.7 pts 50Long-Horizon+27.8 hrs

Long Context

56Long Context+34.0 pts 561M Recall+29.5 pts 51Long Reasoning+28.8 pts

Speed & Cost

45Speed & Cost+48.6 pts 33Output Speed+448 tok/s 17Latency+1.23s 53Cost-Efficiency+75.4 pts

Trust & Factuality

51Trust & Factuality+29.4 pts 52Factuality+24.0 pts 51Honesty & Calibration+24.3 pts

Overall Intelligence over time

#1 frontier3.7 SonnetV4.1 FlashGLM-5.3

Update history

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

The Claude Sonnet family

3.7 SonnetFebruary 2025 Sonnet 4.5September 2025 Sonnet 5.5September 2026

60-second briefing

3.7 Sonnet, 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

3.7 Sonnet is 52nd in overall intelligence; leader is Claude Opus 5.5.