Anthropic released Claude Sonnet 5.5 on 2026-09-28 (API id claude-sonnet-5-5), the second Claude 5.5-family model. Vs Sonnet 5: Terminal-Bench 4.0 70.6% (from 10.3%), CursorBench 4.0 55.5%, GDPval-AA near Opus 5.5; 30%+ faster output and up to ~30% lower task cost at the same $2/$10 pricing. Native 1M context; adaptive thinking on by default — use between_tools to disable up-front thinking.

Key Takeaways

  • ✓Shipped 2026-09-28 — model id claude-sonnet-5-5 (announcement)
  • ✓Terminal-Bench 4.0 70.6% (vs Sonnet 5 10.3%); CursorBench 4.0 55.5%; GDPval-AA 1844
  • ✓Same $2/$10 pricing; ~30%+ faster output; up to ~30% lower task cost via fewer tokens
  • ✓1M context / 128K max out; adaptive thinking default; between_tools turns off up-front thinking
  • ✓Migrate via docs; anthropic-sdk-python ≥1.9.0; Claude Code ≥2.1.284 sonnet alias → 5.5
🧭

Finished reading? Explore benchmark rankings & pricing

Real-world SWE-bench scores & $20/mo vs API cost break-even calculator

🏆 Leaderboard→💰 Compare Plans↗
🔗

Project Links & Resources

Direct Access
Direct access to official project resources and documentation
📝
Introducing Claude Sonnet 5.5 (Anthropic)
www.anthropic.com/claude-sonnet-5-5
↗
📖
Claude Sonnet 5.5 model overview
platform.claude.com/docs/en/models/sonnet-5-5/overview
↗
📖
Migration guide
platform.claude.com/docs/en/models/sonnet-5-5/migration-guide
↗
📖
What's new in Claude Sonnet 5.5
platform.claude.com/docs/en/models/sonnet-5-5/whats-new-sonnet-5-5
↗
🌐
Claude Sonnet 5.5 System Card
www.anthropic.com/claude-sonnet-5-5-system-card
↗
🐙
anthropic-sdk-python v1.9.0
github.com/anthropics/anthropic-sdk-python/releases/tag/v1.9.0
↗
🔬

In-Depth Technical Analysis

Core Background


Claude Sonnet 5.5 (released 2026-09-28) is the second Claude 5.5-family model: a faster, lower-cost complement to Opus 5.5 for well-scoped coding, bugfix, and polished docs/slides/spreadsheets. Official post: https://www.anthropic.com/claude-sonnet-5-5

Architecture


Model ID claude-sonnet-5-5 (Bedrock anthropic.claude-sonnet-5-5). Native 1M context, 128K max output ($2/$10 per MTok, cache read $0.20). Adaptive thinking on by default; use between_tools to turn off up-front thinking. First Sonnet with Opus-class cyber safeguards and distillation classifiers.

Benchmarks (official)


Terminal-Bench 4.0: 70.6% vs Sonnet 5 10.3%. CursorBench 4.0: 55.5% (Opus 5.5 57.8%). GDPval-AA v2.1: 1844 vs 1449. OSWorld 2.1 partial: 80.1% vs 57.0%. Output ~30%+ faster; up to ~30% lower task cost via fewer tokens. See System Card.

Developer guide


Pass claude-sonnet-5-5; migrate thinking-off → between_tools. Claude Code ≥2.1.284 sonnet alias → 5.5. Upgrade anthropic-sdk-python ≥1.9.0. Review five breaking changes in the migration / what's-new docs before cutover.