← All models

Claude Sonnet 5

Published Publisher Anthropicclaude-sonnet-5

Claude Sonnet 5

Positioning and capabilities

Claude Sonnet 5 is an agent-oriented model in Anthropic’s Sonnet family, released on June 30, 2026. Anthropic describes it as an improvement over Sonnet 4.6 in reasoning, tool use, coding, and knowledge work. It can plan multi-step tasks, use tools such as browsers and terminals, and operate autonomously. At release, Anthropic reported performance close to Opus 4.8 on some tasks at a lower price. (anthropic.com)

It accepts text and images and generates text. Its context window is 1 million tokens, with up to 128,000 output tokens. Adaptive thinking is enabled by default; the default effort level is high. (platform.claude.com)

Uses

Potential applications include agentic software development and debugging, browser or computer-use workflows, research, knowledge work, and multi-step business automation. Tool-enabled workflows require the relevant tools and integration. (anthropic.com)

Limitations and status

  • The training-data and reliable-knowledge cutoffs are January 2026. Outputs and autonomous actions should be checked, particularly for consequential tasks. (platform.claude.com)
  • The Message Batches API supports up to 300,000 output tokens only with the output-300k-2026-03-24 beta header. (platform.claude.com)
  • Manual extended thinking is not supported; non-default temperature, top_p, and top_k settings return an error. Priority Tier is unavailable. Its tokenizer produces approximately 30% more tokens for the same text than the preceding model, depending on the workload. (platform.claude.com)
  • As checked on November 16, 2026, the model is active but classified as legacy; Anthropic says retirement will be no earlier than June 30, 2027. (platform.claude.com)

Claude API pricing

The following Anthropic list prices are in USD per million tokens, apply to all context-window sizes, and were checked on November 16, 2026. They are Claude API prices, not necessarily prices for other platforms. Cache write and read columns refer to cached input tokens. (assets.claude.com)

Processing and inference scope Base input Output 5-minute cache write 1-hour cache write Cache read/refresh
Standard, global $2.00 $10.00 $2.50 $4.00 $0.20
Standard, US-only inference $2.20 $11.00 $2.75 $4.40 $0.22
Batch, global $1.00 $5.00 $1.25 $2.00 $0.10
Batch, US-only inference $1.10 $5.50 $1.375 $2.20 $0.11

The global standard input/output prices were initially introductory prices and became standard on August 10, 2026; the previously scheduled increase was canceled. (platform.claude.com)

Official sources