Anthropic has released Claude Opus 5, maintaining the same pricing as Opus 4.8 while significantly improving performance on agentic coding benchmarks. This new model offers capabilities close to Fable 5 at a substantially lower cost, potentially making it the default choice for complex tasks. However, the release includes two breaking API changes: adaptive thinking is now enabled by default, which may truncate responses, and specific configurations for disabled thinking will result in errors. AI
IMPACT Sets a new price-performance standard, potentially reducing the need to escalate to premium models for complex tasks.
RANK_REASON Frontier-lab model release with system card and benchmark data. [lever_c_demoted from frontier_release: ic=1 ai=1.0]
- Amazon Bedrock
- Anthropic
- Claude API
- Claude Opus 5
- CursorBench 3.2
- Frontier-Bench v0.1
- Google.Cloud
- Microsoft Foundry
- Opus 4.8
- OSWorld 2.0
AI-generated summary · Google Gemini · from 1 sources. How we write summaries →
