Platform engineering teams
Standardize agent environments, permissions, models, secrets, and execution across multiple repositories.
Independent tool overview
Oz at a glance
Oz is Warp's orchestration platform for running, tracking, automating, and steering coding agents locally or across hosted and self-hosted cloud environments.
Visit the official Oz site ↗
- Product
- Coding-agent orchestration platform from Warp
- Run modes
- Local, Warp-hosted cloud, and Enterprise self-hosted workers
- Interfaces
- Warp app, web dashboard, Oz CLI, API, and SDK
- Triggers
- GitHub, Slack, Linear, schedules, and custom webhooks
- Pricing model
- Plan fee plus usage-based agent credits
- Last reviewed
- August 28, 2026
Overview
What Oz is
Oz is the agent platform behind Warp's local and cloud coding workflows. It is built for developers and platform teams that want to run agents in parallel, trigger them from engineering systems, and inspect what every run did.
The product supports interactive local work as well as background cloud agents. Cloud runs can react to GitHub, Slack, Linear, schedules, and custom webhooks, while the web app, CLI, API, and SDK provide different ways to launch and manage the same underlying agent workflows.
A configured cloud environment defines the Docker image, repositories, setup commands, secrets, and runtime used for a task. Teams can use Warp-hosted workers, and Enterprise customers can run self-hosted workers inside their own infrastructure.
Oz is most valuable when a team already knows which software tasks are safe and repeatable enough to automate. It supplies orchestration and auditability, but it does not remove the need for repository permissions, spend controls, tests, review gates, and ownership of agent-created changes.
Use cases
Who Oz is best for
The strongest fit depends on the job you need the product to complete, not the size of its feature list.
High-volume maintenance
Run repeatable work such as dependency updates, code cleanup, issue triage, and PR review in the background.
Teams coordinating parallel agents
Dispatch many coding tasks at once while retaining a dashboard, audit trail, and the ability to steer individual runs.
Agent-powered internal products
Use the API, SDK, CLI, and webhooks to build custom engineering automations on top of cloud agents.
Capabilities
Core Oz features
Local and cloud agents
Use the same Warp agent capabilities interactively on a developer machine or asynchronously in a remote environment.
Parallel orchestration
Launch multiple agents across repositories and tasks rather than waiting for one local session to finish.
Event triggers and schedules
Start runs from GitHub, Slack, Linear, custom webhooks, or recurring schedules for unattended workflows.
Web dashboard and audit trail
Browse runs, inspect transcripts and actions, share sessions, and monitor agent activity from a desktop or mobile browser.
CLI, API, and SDK
Launch and manage agents from a terminal, CI pipeline, backend service, script, or custom application.
Reusable environments
Define a Docker image, repositories, setup commands, runtime settings, and secrets once and reuse them across triggers.
Multi-model and skill support
Choose from Warp's available models, run reusable Skills, and connect MCP servers for access to systems such as GitHub, Linear, and Sentry.
Self-hosted execution
Enterprise customers can keep agent workloads inside their own VPC or on-premises environment while Oz provides the control plane.
Process
How the Oz workflow works
Step 1
Define the automation
Choose a narrow software task with clear inputs, expected outputs, tests, and an owner for exceptions.
Step 2
Configure an environment
Select the container image, repositories, setup commands, secrets, permissions, model, and execution host.
Step 3
Choose how it starts
Run the task manually or connect a GitHub, Slack, Linear, schedule, API, CLI, or webhook trigger.
Step 4
Monitor and steer
Follow the transcript and actions in Warp or the Oz web app, and pause or redirect the run when necessary.
Step 5
Review the output
Require automated checks and human review before merging code, changing infrastructure, or acting on production systems.
Cost
Oz pricing and free plan
Oz is included across Warp plans, but agent runs consume credits based on AI usage and, for Warp-hosted cloud runs, compute. Additional usage is pay as you go. Annual prices below reflect Warp's listed 10% discount.
Free
$0/month
Core terminal and limited cloud-agent access for individual evaluation.
- Warp Agent CLI access
- Limited cloud agents and cloud conversation storage
- Bring-your-own inference option
- Additional agent usage available at pay-as-you-go rates
Build
$20/month or $18/month annually
Usage-based individual plan with broader model and cloud-agent access.
- 1,500 credits, described by Warp as $20 of included agent usage at API rates
- Extended cloud-agent access
- Unlimited cloud conversation storage
- Reload credits with usage controls and volume-based discounts
Max
$200/month or $180/month annually
Higher-capacity individual plan for frequent agent workloads.
- 18,000 included credits
- Includes the Build feature set
- Additional usage remains pay as you go
Business
$50/user/month or $45/user/month annually
Team plan for up to 25 seats with administration and security controls.
- 1,500 credits per seat
- Team usage metrics and spend controls
- Admin-configurable data controls and SAML SSO
- Bring your own API keys and custom inference endpoints
Enterprise
Custom
Organization plan for advanced governance, deployment, and support.
- Unlimited seats and custom shared credit pools
- Self-hosted cloud agents and bring-your-own-LLM routing
- Advanced controls, analytics, and codebase indexing
- Dedicated implementation and account support
Pricing checked . Check current pricing at the source ↗
Assessment
Oz strengths and limitations
Where it stands out
- Combines interactive local agents and background cloud automation in one platform.
- Supports many launch surfaces, including a web app, CLI, API, SDK, integrations, schedules, and webhooks.
- Provides environment standardization, run history, shareable sessions, and in-flight steering.
- Multi-model architecture reduces dependence on a single model vendor.
- Enterprise self-hosting can keep execution and source code inside a customer's network.
What to consider
- Usage-based credits combine model and hosted-compute costs, so high-volume automation needs active spend monitoring.
- Cloud agents can make broad code and system changes; safe deployment requires tight repository scope, permissions, tests, and review gates.
- Some workflows require team membership, and self-hosted cloud execution is reserved for Enterprise.
- Cloud conversations and Oz runs require conversation data to be stored for the product to function, even though Warp's contracted model providers operate under zero-data-retention terms.
- A standardized Docker environment takes setup and maintenance before automated runs are dependable across a complex codebase.
Compare
Oz alternatives
The right alternative depends on the specific output, workflow, controls and budget your project requires.
Coding
Codex
Choose Codex for OpenAI's coding agent across local, cloud, and app-based workflows when orchestration infrastructure is not the primary requirement.
Explore Codex →Coding
Replit Agent 4
Choose Replit Agent 4 for a browser-first multi-agent product focused on building and iterating applications.
Explore Replit Agent 4 →Coding
Devin Desktop
Choose Devin Desktop for Cognition's integrated workspace for supervising Devin and other coding agents.
Explore Devin Desktop →Questions
Oz FAQs
What is Warp Oz?
Oz is Warp's orchestration platform for running, automating, tracking, and steering coding agents locally or in cloud environments.
Can Oz run agents in parallel?
Yes. Parallel cloud-agent execution is one of Oz's core uses, allowing teams to dispatch many tasks across repositories or workflows at the same time.
What can trigger an Oz agent?
Oz supports manual runs plus triggers from GitHub, Slack, Linear, schedules, custom webhooks, the CLI, API, and SDK.
Can Oz run on my own infrastructure?
Yes, but managed self-hosted workers are an Enterprise capability. Warp says the execution plane can stay inside the customer's infrastructure while its hosted control plane coordinates the runs.
How much does Oz cost?
Oz access begins on Warp's Free plan. Paid plans start with Build at $20 monthly, or $18 monthly on annual billing, and runs consume credits based on model usage and hosted compute.
Does Warp train models on Oz data?
Warp says its contracted LLM providers operate under zero-data-retention agreements and do not train on customer data. Oz cloud runs still store conversation data needed for run history and the product experience, so teams should configure their privacy and retention controls deliberately.
Bottom line
Our Oz verdict
Warp Oz is a capable orchestration layer for engineering teams ready to move beyond one-off coding-agent sessions into scheduled, event-driven, parallel automation. Its mix of interfaces, auditability, and hosting options is strong, but the operational value depends on disciplined environment setup, permission boundaries, review gates, and usage controls.
Visit Oz website ↗Similar Coding Tools
View all Coding AI tools →
Composer 1.5
Composer 1.5 - Cursor's updated in-house agentic coding model with increased performance and speed
Orchids 1.0
Orchids 1.0 - AI app builder for any stack with BYO model subscriptions
GPT-5.3-Codex
GPT-5.3-Codex - OpenAI's new SOTA agentic coding model
M2.5
M2.5 - Minimax's new open-source frontier model with powerful coding capabilities
AI training for the future of work.
Get access to all our AI courses, hundreds of real-world AI use cases, live expert-led workshops, an exclusive network of AI early adopters, and more.
AI Courses
Get unlimited access to all of our current & upcoming industry-specific AI courses for the duration of your subscription.
Daily Guides
To keep up with the rapid pace of AI, our team publishes AI implementation guides daily. Our library contains 300+ practical use cases to automate real-world work.
Workshops
Join weekly, live, interactive sessions with industry leaders who are at the forefront of AI for hands-on implementation guidance and exclusive insights.
Community
Network with an exclusive community of AI-first professionals who are working smarter with AI. Learn how early adopters are using AI in their work and businesses.