Claude Sonnet 4.5 at a glance
Claude Sonnet 4.5 reported 77.2% on SWE-bench Verified and 61.4% on OSWorld, stayed focused for over 30 hours on complex tasks, kept $3/$15 pricing and arrived with the Claude Agent SDK.
What changed with Claude Sonnet 4.5
On September 29, 2025, Anthropic released Claude Sonnet 4.5. It reported state-of-the-art 77.2% on SWE-bench Verified and a leading 61.4% on OSWorld for computer use, up from Sonnet 4's 42.2% four months earlier, and said it had observed the model staying focused for more than 30 hours on complex multi-step tasks.
Pricing stayed at $3 per million input tokens and $15 per million output tokens. Anthropic called it its most aligned frontier model so far and released it under ASL-3 protections. Alongside it came checkpoints in Claude Code and the Claude Agent SDK, which exposes the infrastructure behind Claude Code so developers can build their own agents.
Claude Sonnet 4.5 benchmarks and pricing
Values are copied from the linked source on the date shown. They are not IA Compa measurements.
- Input price per 1M tokensas of · Source for Input price per 1M tokens: anthropic.com (opens in a new tab)
- $3.00
- Output price per 1M tokensas of · Source for Output price per 1M tokens: anthropic.com (opens in a new tab)
- $15.00
How it compares
Other models in the same category that we have covered, using each model's latest recorded value. (Models)
Where to compare Claude Sonnet 4.5
Independent leaderboards and official resources for comparing this model yourself. Rankings change often; each site shows its own current data.
General leaderboards
Official resources
Why Claude Sonnet 4.5 matters for businesses
For businesses, Sonnet 4.5 offered top-tier coding and computer-use ability at a mid-tier price, and the Agent SDK made it far easier to build custom agents for operations, research or support on the same foundation Anthropic uses.
Put it to work: AI Agents
Sources for this story
- Introducing Claude Sonnet 4.5 (opens in a new tab)Official
Anthropic · anthropic.com
- Claude 4.5 Sonnet model page (opens in a new tab)Benchmark
Artificial Analysis · artificialanalysis.ai
Last verified: . IA Compa summarizes public announcements for business readers. We are not affiliated with the companies mentioned. Always confirm details, pricing, and availability with the official source.
- #llm
- #anthropic
- #claude
- #coding
- #computer-use
- #agents
