All Technologies
OpenAI
One of two default LLM providers, picked per use case.
We use OpenAI models where they are the right fit for the task — not by default, and not exclusively. Model choice is a per-feature decision based on latency, cost, and eval performance on the specific task, not a blanket vendor commitment.
Where we use it
- LLM features where GPT models score best on our task-specific eval harness
- Function calling and structured output for agentic workflows
- Embeddings for semantic search, benchmarked against alternatives per project
Relevant services
Related reading
Building something with OpenAI?
Start a project