Claude Sonnet 5.5
Faster, cheaper mid-tier model for everyday tasks and agentic work
Context window
1M
Input / 1M tokens
$2.00
Output / 1M tokens
$10.00
Provider
Anthropic
Cache reads $0.10 per 1M (half of Sonnet 5's $0.20) and 5-minute cache writes $2.50 per 1M; Batch API is 50% off. The full 1M context is billed at standard rates. · Data verified 2026-11-08
Claude Sonnet 5.5 (September 28, 2026) is Anthropic's current Sonnet, built for everyday coding, agent, and enterprise work such as code generation, data analysis, documents and agentic tool use. Anthropic says it generates output more than 30% faster than Sonnet 5 and costs up to 30% less per task at the same $2/$10 list price, with cache reads cut to $0.10 per 1M. It scores close to Opus 5.5 on many benchmarks and beats it on Terminal-Bench 4.0, though Anthropic says Opus 5.5 remains stronger at complex, open-ended work.
Capability index
Relative estimates (0-100) to place this model against its peers, grounded in published benchmarks.
How to access it
Available via the Claude API (claude-sonnet-5-5), Amazon Bedrock (anthropic.claude-sonnet-5-5), Claude Platform on AWS, Google Cloud, and Microsoft Foundry, and in the Claude apps and Claude Code.
Strengths
- ✓More than 30% faster output than Sonnet 5 (Anthropic-reported)
- ✓Strong agentic coding performance
- ✓Excellent at document generation and formatting
- ✓Design-oriented capabilities
- ✓Same $2/$10 price as Sonnet 5, with cache reads halved to $0.10 per 1M
Best for developers who...
When to choose it (and when not to)
Reach for Claude Sonnet 5.5 when...
- →Need faster responses than Sonnet 5
- →Working with well-scoped tasks
- →Cost-conscious deployment at scale
- →Agentic applications
Look elsewhere if...
- ✕Complex reasoning requiring sustained judgment
- ✕Open-ended problems needing careful analysis
- ✕Tasks where Opus 5.5 is necessary
How to use it
- ›Thinking type disabled returns a 400; send thinking type between_tools (at effort high or below) to turn off up-front thinking
- ›Re-tune effort: levels were recalibrated from Sonnet 5, and high is the API default
- ›Structure prompts around well-defined task boundaries
- ›Remove temperature, top_p and top_k, which return a 400
Quickstart
Pythonfrom anthropic import Anthropic
client = Anthropic()
message = client.messages.create(
model="claude-sonnet-5-5",
max_tokens=1024,
messages=[{"role": "user", "content": "Hello!"}]
)
print(message.content[0].text)Use model ID 'claude-sonnet-5-5' with Anthropic SDK
API model id: claude-sonnet-5-5
Benchmarks
| Benchmark | Score | Notes |
|---|---|---|
| Terminal-Bench 4.0 | 70.6% (max effort) | Anthropic-reported; Sonnet 5 scored 10.3% |
| CursorBench 4.0 | 55.5% | vs 34.1% for Sonnet 5 |
| GDPval-AA v2.1 (Knowledge Work) | 1844 Elo | Two points behind Opus 5.5's 1846 |
| SWE-bench Pro | 81.3% (max effort) | Anthropic system card, average of 5 trials |
| DeepSWE v1.1 | 71.0% (max effort) | Anthropic system card, average of 5 trials |
| OSWorld 2.1 | 80.1% (max effort) | Anthropic-reported, full 108-task set, partial score |
Source: Anthropic official announcement
Compare Claude Sonnet 5.5
Claude Sonnet 5.5 vs GPT-6.1 Sol
OpenAI - 1,050,000 ctx
Claude Sonnet 5.5 vs Kimi K3.1 Pro
Moonshot AI - 1M ctx
Compare Claude Sonnet 5.5 with any other model
Build a comparison →All model comparisons →