We've reimagined agent orchestration UX with our new Warp Agent CLI, making it easier than ever to see your subagents. Here's Moira, one of our engineers, walking through some of our features! Configure the model, harness, and environment your subagents run with all with a couple shortcuts

Try it in Warp:

Claude Opus 5 is live in Warp, and we're seeing a 43% reduction in cost per task at nearly the same quality as Fable 5 in our agent evals.

Finally, you can drag tabs into and out of the terminal window 🙌

The Warp Agent now supports Claude Sonnet 5. We're seeing significant gains in agentic coding and professional work, all at Sonnet pricing. A good default to set for tasks around your terminal!

We're here at AI Engineer World's Fair! Stop by our booth to chat with the team and our CEO Zach Lloyd

You can now connect your SuperGrok subscription directly to the Warp Agent. Get access to the latest models, including Grok Build 0.1.

Warp now has tab groups. Group together your work on related projects without juggling separate windows or split panes. Try it on the latest preview build.

Warp now supports Claude Fable 5. Fable 5 has Mythos-level performance and is highly capable of /goal-oriented, unsupervised tasks. Ready to embed into your next loop 🔁

Warp now generates a commit message for you. Open the code review panel, click "commit," and tweak the message to your liking.

The Warp Agent now supports message queueing: - /queue to queue the next message - Toggle ⌘⇧J to auto-queue messages for the session - Drag to re-order messages in the queue - Send messages right away as-needed It's all the options you could need and more!

Introducing Agent Memory: shared context that helps agents remember what works across conversations. ✅ Cross-harness: Claude Code, Codex, Warp Agent ✅ Cross-team: learns across everyone’s sessions ✅ Portable: turnkey hosting or self-hosted Now in research preview

Warp just crossed 800K monthly developers and over 60K GitHub stars. Thank you all for the support! ⭐

Proud to work with Anthropic to make Claude so accessible. In Warp, you all have: Run 40 million Warp Agent conversations Started 10 million Claude Code sessions Used 55 *billion* Claude tokens per day Thank you for the support!

Warp now supports the file tree and code review over SSH connections. We've also added full coding capabilities to the Warp Agent over SSH (namely file editing and codebase indexing). Just install the SSH extension the next time you start a session

Warp now has buttons to commit, push, and PR the code your agents write. Open the code review panel to try it out.

Warp also supports Claude Opus 4.8 Fast, with Opus-class performance at at 3x lower cost compared to Opus 4.7 Fast

Warp now supports Claude Opus 4.8. Opus 4.8 is built for longer agent runs: planning up front, following through across multi-step coding tasks, and catching real bugs in review instead of stopping after one clean test run.

You can now connect your @openrouter.bsky.social URL directly to the Warp Agent. OpenRouter unlocks a world of models including Anthropic, OpenAI, Gemini, DeepSeek, Kimi, and more, all with better prices and uptime. Set your URL, model, and an alias to easily switch using /model.

The Warp Agent now supports Gemini 3.5 Flash. It's an impressive combination of intelligence, speed, and cost.

We've added local-to-cloud handoff for all Warp Agent conversations. Type & to hand off agents from your terminal to the cloud. Close your laptop, let them run overnight, and pick them up locally in the morning.

We are excited to introduce Agent Memory in research preview. Agent Memory is a cross-harness memory system that helps Claude Code, Codex, and the Warp agent remember what works across every session. Join the waitlist: warp.dev/oz/agent-me... 🔖

Oz can now orchestrate subagents automatically by deploying multiple agents in parallel. Each agent gets its own worktree when run locally, and its own isolated Docker container in the cloud. Use /orchestrate or let the Warp Agent suggest a delegation plan.

You shouldn’t have to bet your future on a single model or harness. Oz now gives you the ability to launch, track, and steer Claude Code, Codex, and the Warp Agent from the same cloud platform. Multi-harness orchestration is available to all users while in beta.

Warp now renders markdown tables, mermaid diagrams, and images in all agent responses. No more stray Markdown formatting to read 👏

The Warp team uses a lot of Skills to speed up our workflow, so we open-sourced them. > npx skills add warpdotdev/oz-skills Includes Skills for SEO and accessibility auditing, documentation writing, DevOps tasks like terraform configuration, and more.

Warp's settings are now fully agent-accessible from a settings[.toml] file. Just point your favorite coding agent at your Warp configuration directory and describe the changes you want to make.

Warp's agent now has an `auto (open-weights)` model. This will route you to frontier-level open weight models based on the complexity of your task. High performance at a reduced token cost compared to "genius" mode.

After open-sourcing the terminal, we started asking... what else should we open-source? Well, now we've open-sourced our docs. Docs lead Hong Yi wrote about how we migrated from a CMS to a GitHub repo of Markdown files with @astro.build. We built the proof-of-concept in just 3 hours.

Have a bug report or feature you want to see in Warp? Just type /feedback in Warp. This skill lets you work with an agent to post a high-quality issue on our open-source repo. From there, we invite you to contribute a fix!