✸

Claude

Best for coding

by Anthropic

4.8
Benchmark RatingLaunched 2023Free · $20/mo Pro

Claude leads on SWE-bench (real software engineering) and is the top pick for careful, high-quality writing and long documents. Artifacts, Projects and computer use make it a favourite for codebases and deliberate reasoning.

SWE-bench #1ArtifactsProjects
Visit Site

Pricing Tiers

Click a tier to see full details

Free — $0

Subscribe

Sonnet, daily limits

Plan: FreePrice: $0Tool: Claude

Capabilities & Specs

Best for
Long documents & nuance
Context window
200K
Image generation
Voice mode
Code execution
Web browsing
Offline / open weights
Public API
Starting price
Free · $20/mo

Technical Benchmarks

Raw scores from public leaderboards (LMArena, OpenAI, Anthropic). Higher is better.

MMLU89.3%

Language Understanding

HumanEval93.7%

Code Generation

SWE-bench49%

Software Engineering

GSM8K96.4%

Math Reasoning

GPQA59.4%

Graduate QA

IFEval89.3%

Instruction Following

LMArena ELO1271

Chatbot Arena

Pros

  • Best-in-class long-context handling
  • Superior writing quality
  • Artifacts for live previews

Cons

  • No native image generation
  • Tighter content guardrails

Available Models

ModelContextPriceNote
Claude Opus 4.6200K$20/moLatest flagship (Aug 2026)
Claude Sonnet 4.6200KFree · $20/moBalanced default
Claude Opus 4.1200K$20/moPrevious Opus
Claude Sonnet 4.5200KFree · $20/moPrevious Sonnet
Claude Haiku 4.5200KFreeFast + cheap
Claude Fable 5200K$20/moCreative writing tuned
Claude Mythos 5200KLimitedStorytelling & worldbuilding

Tasks Claude Covers

WritingCodingAgents