Warp Terminal Goes Open Source with an AI-First Contribution Model
Warp open-sources its terminal with an AI-first contribution model where agents write code and humans handle ideas and review, plus new model support and licensing details.
Warp, the modern terminal and agentic development environment built in Rust, has made a landmark move by open-sourcing its client code. But this isn't your typical open-source project — the company is introducing an unprecedented contribution model where AI agents do most of the coding while humans focus on ideas and review. Below, we break down the key details of this announcement in a Q&A format.
What exactly is Warp, and what does it mean that it's now open source?
Warp is a next-generation terminal that runs on Linux, Windows, and macOS. It features a block-based command interface and built-in support for AI coding agents like Claude Code, Codex, and Gemini CLI. The company behind it, Warp, has released the client codebase on GitHub under a split license: the UI framework (warpui_core and warpui crates) is MIT-licensed, while the rest of the code is under AGPLv3. This open-sourcing makes the code available for community inspection and contribution, but with a twist — contributions are expected to be driven primarily by AI agents rather than human-written code.

Why did Warp decide to open-source its terminal?
CEO Zach Lloyd explained that the decision stems from a desire to build a successful business. Warp competes with highly funded, closed-source rivals, and opening the project allows the community to help accelerate product development. The company believes that the main bottleneck is no longer writing code but human-led tasks such as deciding on features and verifying software behavior. By open-sourcing, Warp hopes to tap into the collective intelligence of the community while shifting the implementation load to AI agents.
How does the new open-source contribution model work?
Unlike traditional open-source projects where contributors submit pull requests with hand-written code, Warp's model envisions human contributors focusing on ideas, specifications, and code review. The actual implementation is handled by AI coding agents — specifically, Oz, Warp's cloud agent orchestration platform. The company is confident that Oz-generated code, guided by their rules and verification processes, can reliably implement features. This approach is designed to streamline development and let humans concentrate on the higher-level tasks that require creativity and judgment.
What is Oz, and why is it central to the workflow?
Oz is Warp's cloud agent orchestration platform, announced earlier this year. It allows developers to run multiple coding agents in parallel in the cloud, with full visibility and control over their actions. In the open-source contribution workflow, Oz is the preferred agent because it already has the right context and built-in checks for Warp's development process. While other coding agents like Claude Code or Codex are welcome, using Oz ensures smoother integration with the project's requirements. Human contributors can direct Oz to implement specific features, then review the generated code.

What are the licensing terms for the open-sourced code?
The Warp codebase is split into two licensing tiers. The UI framework, which includes the warpui_core and warpui crates, is released under the permissive MIT license. The remainder of the codebase, including the core terminal functionality, is licensed under the GNU Affero General Public License v3 (AGPLv3). This split allows the UI components to be freely used in other projects, while the AGPLv3 ensures that any modifications to the core software remain open source when distributed over a network.
Who is backing this release, and what AI models are supported?
OpenAI is the founding sponsor of the Warp open-source repository, and the agentic contribution workflows are powered by GPT models. However, Warp is also expanding support for other open-source models. The announcement includes the addition of Kimi, MiniMax, and Qwen as options for AI agents. Furthermore, a new routing option called "auto (open)" will automatically select the best open model for a given task. This ensures that contributors have flexibility while still benefiting from the performance of leading models.
What other new features come with this open-source release?
Beyond the code release and model additions, Warp is shipping a settings file that enables programmatic control and easier portability of configurations across devices. This makes it simpler for users to customize their terminal environment and share settings. The move towards open-source model support also signals Warp's commitment to a broader AI ecosystem, allowing users to choose from multiple providers rather than being locked into a single vendor. These updates aim to make Warp more adaptable and community-driven while maintaining its performance and agentic capabilities.