Mythos

Claude Sonnet is 📝Anthropic's mid-tier model line in the 📝Claude family, positioned as the best balance of speed, intelligence, and cost; the current model, Claude Sonnet 5.5, was released September 28, 2026.

Claude Sonnet sits between 📝Claude Opus and 📝Claude Haiku in Anthropic's lineup and is the tier most developers use for everyday production work. Several Sonnet releases have outperformed the Opus model of their day: Claude 3.5 Sonnet beat Claude 3 Opus in 2024, and Anthropic reports that Sonnet 5.5 nearly matches Claude Opus 5.5 on knowledge work at half the per-token price. The tier has also introduced several of Anthropic's defining capabilities, including computer use and extended thinking.

Key Facts

  • Current model: Claude Sonnet 5.5 (claude-sonnet-5-5), released September 28, 2026
  • Pricing: $2 input / $10 output per million tokens; cached reads $0.20
  • Context window: 1M tokens, 128K max output
  • Thinking: adaptive, on by default, with a default effort of high
  • Knowledge cutoff: June 2026
  • Tier position: below Claude Opus and the Mythos tier, above Claude Haiku
  • Status as of: September 28, 2026

Model Lineage

  • Claude 3 Sonnet (March 4, 2024) — mid-tier model of the first three-tier Claude family; retired July 2025
  • Claude 3.5 Sonnet (June 20, 2024) — outperformed Claude 3 Opus at Sonnet pricing and launched alongside Artifacts
  • Claude 3.5 Sonnet, upgraded (October 22, 2024) — introduced computer use in public beta
  • Claude 3.7 Sonnet (February 24, 2025) — first Claude model with extended thinking; launched with the 📝Claude Code research preview
  • Claude Sonnet 4 (May 22, 2025) — mid-tier half of the Claude 4 launch
  • Claude Sonnet 4.5 (September 29, 2025) — coding- and agent-focused, released with the Claude Agent SDK
  • Claude Sonnet 4.6 (February 17, 2026) — Anthropic's designated replacement for retired Sonnet 3.x and 4 models
  • Claude Sonnet 5 (June 30, 2026) — first Sonnet of the Claude 5 generation
  • Claude Sonnet 5.5 (September 28, 2026) — current; second model of the Claude 5.5 family

Claude Sonnet 5.5 (Current)

Claude Sonnet 5.5 keeps Sonnet 5's $2 / $10 pricing while, according to Anthropic, generating output more than 30% faster and costing up to 30% less per task by using fewer tokens and tool calls. Anthropic reports a GDPval-AA v2.1 score of 1844 against Opus 5.5's 1846, and 70.6% on Terminal-Bench 4.0, ahead of Opus 5.5's 66.4%. Unlike Opus 5.5, its thinking can be reduced to a between_tools setting that skips up-front thinking. Code written for Sonnet 5 must drop forced tool use and, on the Claude API and 📝Google Cloud, move off the computer_20251124 tool.

Why It Matters

Sonnet is where most Claude traffic runs, so each Sonnet release resets the price floor for capable agentic work. With Sonnet 5.5 landing within a few points of Opus 5.5 on knowledge work at half the price, the practical choice for many teams shifts from "which tier is smart enough" to "which tasks justify Opus."

FAQ

What is the latest Claude Sonnet model?

Claude Sonnet 5.5, released September 28, 2026, is the latest Sonnet model. Its API model ID is claude-sonnet-5-5.

How much does Claude Sonnet cost?

Claude Sonnet 5.5 costs $2 per million input tokens and $10 per million output tokens, the same as Sonnet 5, with Batch API requests at half price.

Is Claude Sonnet 5.5 as good as Claude Opus 5.5?

Anthropic reports that Sonnet 5.5 nearly matches Opus 5.5 on knowledge work and beats it on Terminal-Bench 4.0. Anthropic positions Sonnet 5.5 for well-scoped work such as bug fixing, everyday coding, documents, and spreadsheets, and Opus 5.5 for longer, harder agentic runs.

What is the difference between Claude Sonnet and Claude Haiku?

Sonnet is the balanced tier; Haiku is Anthropic's fastest and cheapest. The current Haiku, Claude Haiku 4.5, has a 200K-token context window and a February 2025 knowledge cutoff, while Sonnet 5.5 has a 1M-token window and a June 2026 cutoff.

Related

Contexts

Created with 💜 by One Inc | Copyright 2026