All Technologies
Anthropic
The other default LLM provider, picked per use case.
Claude models are our other default, used where they win on our evals — often for tasks needing longer context handling or more reliable instruction-following. Same rule as OpenAI: the model is chosen per feature, benchmarked against the alternative, not picked on vendor loyalty.
Where we use it
- Long-context tasks where a large context window matters more than raw latency
- Agentic workflows built with Claude's tool-use capabilities
- Tasks where our eval harness shows Claude outperforming the alternative
Relevant services
Related reading
Building something with Anthropic?
Start a project