Skip to content
All systems operational0 AI providers monitored, polled every 2 minutes
Live status

Claude Sonnet 5.5

Mid-tier

by Anthropic

Claude Sonnet 5.5 shipped on September 28, 2026 as claude-sonnet-5-5, the second model in the Claude 5.5 family after Opus 5.5, and the price did not move: $2 per million input tokens and $10 output, the same as Claude Sonnet 5, with cache reads at $0.20, cache writes at $2.50 for five minutes and $4 for an hour, and Batch at half price. What changed is how much work a token buys. Anthropic says it generates output more than 30 percent faster than Sonnet 5 and costs up to 30 percent less per task, because it needs fewer tokens to do the same work rather than because the tokens are cheaper. The context window is 1 million tokens with 128K output, and the default effort on the Claude API is high. The jump on Anthropic's own table is large: 70.6 percent on Terminal-Bench 4.0 against 10.3 for Sonnet 5 and 66.4 for Opus 5.5, 64.5 percent on Humanity's Last Exam with tools, and 80.1 percent on OSWorld 2.1. The independent numbers are lower and still strong. Vals measures Terminal-Bench 4.0 at 64.14 percent on its mini-swe-agent run, one point behind Opus 5.5 at 65.15, or 62.63 if the 4 of 198 attempts served by the Sonnet 5 fallback count as failures. Cognition puts it at 52.1 percent on FrontierCode v1.1 at xhigh effort, falling to 46.2 at max. As of September 30, 2026 it sits second on the Vals Index at 67.04 percent, a fraction ahead of Opus 5.5 at 66.97, and Artificial Analysis scores it 56 on its Intelligence Index, behind Opus 5.5. Anthropic's own guidance is that Opus 5.5 remains clearly stronger on complex, open-ended work. It is the first Sonnet to launch with cyber safeguards, so higher-risk cybersecurity tasks visibly fall back to Sonnet 5. Five API changes will break code written for Sonnet 5, among them that forced tool use returns an error, the computer_20251124 tool is not accepted on the Claude API and Google Cloud, and running with thinking off now means the between_tools setting. It is available on the Claude API, Amazon Bedrock, Google Cloud, Microsoft Foundry, and Claude Platform on AWS. Sonnet 5 is not deprecated and stays at the same rate, and Anthropic says Claude Haiku 5.5 will join the family in the coming weeks.

Input Price

$2.00

per 1M tokens

Output Price

$10.00

per 1M tokens

Context Window

1M

tokens

Released

2026-09

API access

Capabilities

textvisiontool-usecodereasoning

Key Strengths

  • ✓$2/$10 per 1M tokens, unchanged from Sonnet 5, with $0.20 cache reads
  • ✓Terminal-Bench 4.0 at 64.14 percent measured by Vals, one point behind Opus 5.5
  • ✓FrontierCode v1.1 at 52.1 percent at xhigh effort, measured by Cognition
  • ✓Second on the Vals Index at 67.04 percent as of September 30, 2026
  • ✓1M token context window with 128K output
  • ✓More than 30 percent faster output than Sonnet 5, per Anthropic

Best For

  • ▸Everyday agentic coding and bug fixing at mid-tier pricing
  • ▸Documents, slides, and spreadsheets
  • ▸High-volume support and knowledge workflows
  • ▸Computer-use agents

Benchmark Scores

BenchmarkScoreDescription
FrontierCode v1.152.1Agentic coding on frontier software engineering tasks (Main split)
Terminal-Bench 4.064.1Long-horizon agentic work in a terminal across software, science, ML, operations, hardware, security, and media (66 tasks, all-or-nothing verifiers)
Humanity's Last Exam (tools)64.5Multidisciplinary expert-level reasoning with tool access

Scores sourced from public benchmark datasets. See full benchmark leaderboard for all models.

Pricing Details

Input tokens

$2.00

per 1M tokens

Output tokens

$10.00

per 1M tokens

Estimated cost per 1K requests

$7.00

~1K input + ~500 output tokens avg

Prices are subject to change. Check the official documentation for current pricing. See the cost calculator for detailed estimates.

Related Models

View DocumentationCompare ModelsCost CalculatorFull Pricing Guide