Gemini is a powerful option in Oz for teams that want access to Google’s latest models and infrastructure. Its strengths in multimodal understanding and fast iteration make it a compelling choice for certain agent workflows, and we like giving developers the flexibility to pick the right model for the job.
OpenAI models are a core option in Oz because of their strong general-purpose reasoning, reliability, and ecosystem maturity. They’re great default agents for debugging, refactoring, and automation workflows, and they integrate cleanly into the kinds of agent orchestration patterns we’re enabling with Oz.
Claude is one of our go-to models in Oz for long-context reasoning, code understanding, and complex multi-step tasks. It’s especially strong at following instructions and staying coherent across large agent workflows, which makes it a great fit for cloud agents that need to reason over entire repos or long-running tasks.
We use Wispr Flow to power voice mode in Warp because it’s the most natural-feeling speech-to-text experience we’ve tried. Low latency, high accuracy, and it actually fits into a developer workflow without getting in the way. It makes talking to agents feel fluid instead of clunky — which matters a lot when you’re iterating quickly.
OpenAI – o3 is available in Warp today. Deeper chain-of-thought, longer context, and fast response times. Deploy AI agents to setup, code, and debug in Warp using o3.
Claude 4 Sonnet is available in Warp today. Claude 4 Sonnet was instrumental in helping Warp agents achieve a score of 71% on SWE-bench and ranking #1 on terminal-bench.
