Runtm

Open-source sandboxes where coding agents build and deploy. Spin up isolated environments where Claude Code, Cursor, and other agents code and deploy software.

Last verified:

Visit Runtm

What is Runtm?

Runtime (runtm) is a cloud development platform that provides sandboxed environments for AI coding agents like Claude Code, Cursor, Codex, and Gemini CLI to build and deploy software safely. It gives engineering teams control and visibility to enable anyone in their organization to use coding agents securely, with isolated sandboxes, live URLs, logs, and previews. The platform sits between your team and the coding agents they use, working with any AI IDE or CLI.

Key features include isolated cloud sandboxes where agents code with full permissions without touching your machine, instant deployment to live HTTPS endpoints, real-time logs and previews for debugging, and a Mission Control dashboard for monitoring all agent sessions. Runtime supports integrations with Slack, Linear, GitHub, Jira, PagerDuty, Sentry, Datadog, and many other company tools. It offers environment templates, bring-your-own-keys (BYOK), role-based access control, spend controls, and audit logs.

Runtime is designed for engineering leaders rolling out coding agents across their organization, as well as product, design, marketing, support, finance, and people teams who want to automate work without filing engineering tickets. Platform teams use it to set guardrails and secrets, while non-engineers get scoped agents to build prototypes, dashboards, and automations safely.

The platform is bring-your-own-cloud with Fly.io as the default provider (AWS and GCP coming soon), and it is self-hostable for enterprises who need to run Runtime inside their own VPC. It supports TypeScript, Python, Ruby, Java, Go, Rust, Docker, and Kubernetes out of the box.

Runtm pricing

Pricing model: Freemium

Free: $0/month with 500 credits included, no credit card required, 1 session, Claude Code/Codex/OpenCode, base sandbox (4 vCPUs, 4 GB RAM), live preview & terminal, GitHub import/export, community support. Builder: $29/month with 10 parallel sessions, all coding agents, standard sandbox (8 vCPUs, 8 GB RAM), live preview & terminal, GitHub import/export, custom templates, BYOK, SDK/CLI/API access, priority support. Teams: Starting from $99/seat/month with unlimited sessions, all coding agents, standard sandbox, real-time collaboration, shared templates & secrets, team activity dashboard, role-based access control, spend controls & usage limits, BYOK, SDK/CLI/API access, priority support. Enterprise: Custom pricing, everything in Teams plus self-host on your own infrastructure, bring your own sandbox provider, environment templates, custom sandbox resources up to 32 GB RAM, dedicated infrastructure, custom agent configurations, network allowlist per agent, audit logs & compliance, SLA & uptime guarantee, SSO (SAML) & SCIM, dedicated support engineer.

Runtm pros

  • Isolated cloud sandboxes keep agents from touching your local machine
  • Deploy to live HTTPS URLs instantly with one command
  • Works with any coding agent: Claude Code, Cursor, Codex, Gemini CLI, Copilot, Devin
  • Live visibility into every session: tool calls, file changes, chain of thought
  • Real-time collaboration in browser with live preview updating for all collaborators
  • Spend controls with per-team budgets and hard caps enforced by platform
  • Environment templates let you set up repos once for teams to boot in seconds
  • Bring your own keys (BYOK) so you manage billing directly with AI providers
  • Integrates with Slack, Linear, GitHub, Jira, PagerDuty, Sentry, Datadog out of the box
  • Background agents clear backlog by triggering from tickets, alerts, and Slack
  • 24/7 on-call agent for alert triage that opens PRs with fixes
  • Self-hostable on your own AWS, GCP, or Azure with secrets injected at runtime
  • Role-based access control and audit logs for enterprise compliance
  • No credit card required for Free tier to start building immediately
  • Supports all major stacks: TypeScript, Python, Ruby, Java, Go, Rust, Docker, Kubernetes

Runtm cons

  • AWS and GCP cloud provider support not yet available, only Fly.io for now
  • Free tier limited to 1 session and 500 credits per month
  • Base sandbox on Free tier has only 4 vCPUs and 4 GB RAM
  • Builder plan at $29/month may be costly for solo developers on budget
  • Teams plan starts at $99/seat/month which is expensive for larger teams
  • Enterprise features require talking to founder for custom pricing
  • Rate limit of 10 deploys per hour on default guardrails
  • Build timeout limited to 10 minutes which may not suffice for large projects

Frequently asked questions about Runtm

Who is Runtime for?

Runtime is for any company that wants every function to ship work with a coding agent, not just engineering. Platform teams use it to roll out agents safely, while product, design, marketing, support, finance, and people teams use the same sandbox to automate work they already own.

Do we need to switch coding agents to use Runtime?

No. Runtime hosts the sandboxes, templates, and policies, and runs the agent your team already prefers, including Claude Code, Codex, Cursor CLI, and Gemini CLI. Runtime is agent-agnostic so you can swap between agents freely.

How do agents work inside Slack? Do we always have to tag them?

You can tag them, or let them work proactively. Each team gets its own sandboxed agent in Slack with a name you choose, like @runtime-finance, @runtime-support, or @ada, that you @mention with a question or task. You can also point it at a channel so it picks up new messages on its own without anyone tagging it. Either way, it replies in a thread with the result, the source rows it used, the cost of the run, and an Open Session button.

What integrations come out of the box?

Runtime ships with first-class connectors for data warehouses (Snowflake, BigQuery, Redshift), billing (Stripe, NetSuite, QuickBooks), HR (Rippling, Gusto, Workday, Deel), CRM and marketing (HubSpot, Segment, GA4), support (Zendesk, Intercom), alerting (PagerDuty, Sentry, Datadog), and engineering (GitHub, Linear, Notion). You can also bring any tool with an API key through its CLI, MCP server, SDK, or REST API.

Does an agent ever touch our raw production data?

No. Agents work against sandboxes that mirror or sample your data with the policies you set, including PII redaction and row-level scopes. Production writes happen only through reviewed actions or pull requests, so the agent never edits live systems on its own.

Can I self-host Runtime?

Yes. You can run Runtime fully inside your own cloud with your own models, sandboxes, and storage, or use their hosted version. Enterprise customers can run Runtime inside their own AWS, GCP, or Azure account so secrets, source code, and model traffic never leave their network. Policies and audit logs follow your team across both.

What's the difference between Base and Standard sandboxes?

Every session runs in its own isolated cloud sandbox. Base sandboxes (Free) have 4 vCPUs and 4 GB of RAM. Standard sandboxes (Builder and above) have 8 vCPUs and 8 GB, which makes a real difference for larger projects and heavier builds. Enterprise can go up to 32 GB of RAM.

What is BYOK (bring your own keys)?

You can plug in your own API keys for AI providers like Anthropic or OpenAI. You manage your own usage and billing directly with the provider, and Runtime handles the orchestration. This is available on Builder, Teams, and Enterprise plans.

How do we control cost as more teams start using agents?

Mission Control shows live spend by team, by repo, and by model, with per-team budgets and hard caps. Admins on Teams and Enterprise can set cost limits per session, per day, or per month for each user. You set the policy, and the platform enforces it before usage runs ahead of the plan.

Can a non-engineer use Runtime safely?

Yes, and that is the point. Product, design, support, and finance get scoped templates that limit which repos and tools they can use, and every change still flows through your normal pull request review. Sandboxes are isolated by default and only receive the secrets, repos, and scopes you assign.

Categories

Use cases

Browse all AI tools on NeedAnAI