MIOSA MIOSA Docs
DocsLearnAPI referenceSDKsCLI & MCPChangelog
DocsLearnAPI referenceSDKsCLI & MCPChangelog

  • Introduction
    • Overview
    • Sandbox to deployment
    • Agent-driven build
    • Docker Host
  • Install the CLI or SDK
  • Sandboxes or Computers?
  • First production app
  • What you can build

  • Start learning
  • Learn and earn $100
  • Learn the words
  • Learn with games
  • Learn with videos
    • Overview
    • Resource model
    • The publish pipeline
    • Computers (desktop product)
    • Tenancy
    • Tenant configuration
    • Data and bindings
    • Preview and share tokens

    • Overview
    • Project lifecycle
    • Project resources
    • Members and usage
    • Overview
    • Quick start
    • Run modes and types
    • Static sites from Git
    • Settings as code
    • Variables
    • Environments
    • Preview environments
    • Releases and rollback
    • Move from Railway
    • Replicas and scaling
    • Dedicated vCPU
    • Logs
    • Metrics
    • CLI
    • SDKs
    • Overview
    • Hosts
    • Templates
    • Docker Host deployments
    • How it works
    • Publishing
    • Deploy from GitHub
    • Versions
    • Releases
    • Rollback
    • Environments
    • Runtime instances
    • Domains
    • Why Forge
    • Quickstart
    • Pull requests and checks
    • From a sandbox
    • Agents and the API
    • CLI reference
    • Limits and FAQ

    • Overview
    • Capabilities
    • Create and connect
    • Docker in sandboxes
    • Lifecycle
    • Files, exec, and artifacts
    • Sizing and limits
    • Timeouts and auto-stop
    • Sandbox events
    • Connecting to a running sandbox
    • Metadata and labels
    • Long-running tasks
    • Production checklist
  • Machine capabilities
    • Snapshots and persistence
    • Templates
    • Sandbox shares
    • Runtime variables
    • Secrets
    • Connections (OAuth)
    • Managed tools for agents
    • SOMA
    • Choosing a runtime
    • Overview
    • How browser use works
    • Quickstart
    • Live viewer
    • Drive it over CDP
    • Integrations
    • Limits, billing, and security
    • All recipes
    • Agent builds an app
    • Agent with secrets
    • Streaming code execution
    • Browser agent

    • Overview
    • Inside a Computer
    • Computers vs sandbox
    • Fast start and readiness
    • Embedding and streaming
    • Lifecycle and sizing
    • Production checklist
    • Overview and concepts
    • Quickstart
    • API reference
    • SDK examples
    • Limits
    • Computer use vs browser use: which to pick
    • OpenComputers
    • Connect your first host
    • Hosts and capabilities
    • Containers on a host
    • Meshes
    • GitHub Actions runners
    • Backups and snapshots
    • Tunnels and exposing services
    • Devices
    • Security model
    • Billing and limits
    • OpenComputers troubleshooting
    • API, CLI, and SDK reference

  • Overview
    • Postgres
    • MySQL
    • Redis cache
    • Qdrant vectors
    • MongoDB
    • Storage and buckets
    • Volumes
    • Auth as a service
    • Transactional email
    • Fractal Fractal enterprise data

  • Domains
  • Previews
  • Internet access and egress
  • Network access rules
  • Edge middleware
  • Protection and WAF
  • Public database endpoint

    • Operating guide for agents
    • What agents can do
    • Agent CLI workflow
    • Agent safety and status
    • Integrating an AI agent
    • Configure an agent
    • Chat with an agent on a machine
    • Agent Studio
    • Harnesses and auth
    • Agent confinement
    • Agent runtime routing
    • Claude Code skills
    • Agent Development Kit (ADK)
    • Agent runtime environment
    • Agent product model
    • Agent product reference map
    • Build an agent-company platform
    • Workflow patterns
    • Choose an agent device
    • Connect agent accounts
    • Agent device model
    • Agent orchestration
    • Runtime profiles and artifacts
    • Orchestration patterns
    • Orchestration examples

    • What Optimal is
    • Quickstart
    • Chats and modes
    • Brains and agents
    • Build on your machines
    • Approvals
    • Tasks, artifacts and history
    • Team chats and sharing
    • Pricing and usage
    • Isolation and permissions
  • Optimal API
  • Troubleshooting

    • What Optimal Engine is
    • Concepts
    • Quickstart
    • Install and configure
    • Operations
    • Sources and ingestion
    • Memory and retrieval
    • Agents and workflows
    • API reference
    • CLI, SDKs and MCP
  • Optimal Engine on MIOSA
  • Troubleshooting

    • Overview
    • Platform builder guide
    • Patterns
    • Agents in your product
    • Templates and sizing
    • Backend orchestration
    • Example applications
    • Build an internal Business OS
    • Ownership and attribution
    • Browser access tokens
    • End-user credentials and OAuth
    • White-label setup
    • Webhooks
    • Agent credit metering
    • All recipes
    • Multi-tenant SaaS
    • Per-user quotas and billing
    • Embed a sandbox iframe
    • Realtime sandbox events
    • Build a virtual device product
    • Overview
    • Foundation UI library
    • Foundation packages
    • React
    • Vue
    • Svelte

  • Build an app like...
    • Lovable Lovable
    • Bolt Bolt
    • v0 v0
    • Replit Replit Agent
    • Base44 Base44
    • Emergent Emergent
    • Rork Rork
    • Anything Anything
    • Same Same.new
    • MIOSA Run agents in the cloud
    • Claude Code
    • Codex
    • Cursor
    • Windsurf Windsurf
    • GitHub Copilot GitHub Copilot
    • OpenHands OpenHands
    • Cline Cline
    • MIOSA Build a coding-agent product
    • Manus Manus
    • Genspark Genspark
    • Devin Devin
    • Claude Cowork
    • Build your own browser agent
    • Polar Polar
    • Browser Use Browser Use
    • Stagehand Stagehand
    • dots dots
    • Grok Bot Grok Bot
    • Meta Meta Muse
    • Characterai Character.AI
    • Replika Replika
    • Pi Pi
    • OpenClaw OpenClaw
    • Perplexity Perplexity
    • Perplexity Perplexity Comet
    • You You.com search and answer API
    • Phind
    • ChatGPT deep research
    • OpenSwarm OpenSwarm
    • Hermes Hermes Agent
    • Delegating orchestrator
    • Elevenlabs ElevenLabs
    • Vapi Vapi
    • Retell Retell AI
    • Bland Bland AI
    • Pipecat Pipecat
    • LiveKit LiveKit Agents
    • Figma Prompt-to-UI tool
    • Uizard Sketch-to-mockup tool
    • Magic Patterns Magic Patterns
    • Runway Video generation
    • Text to video
    • Heygen Avatar video
    • Higgsfield Higgsfield
    • Gamma Gamma
    • Descript Descript
    • Intercom Intercom Fin
    • Decagon Decagon
    • Clay Clay
    • Gong Post-call analysis
    • OpenEvidence OpenEvidence
    • MIOSA Internal Business OS
    • Internal AI workspace
    • Ops and support copilot
    • Internal tools builder
    • Knowledge-base agent
    • Employee onboarding agent

  • CLI or MCP: which to use
    • CLI overview
    • Install the CLI or SDK
    • Account and environment
    • Machines and lifecycle
    • Run commands and move files
    • Snapshots, checkpoints and restore
    • Environments
    • Agents, chats and runs
    • Platform resources
    • Access, org and billing
    • Raw API access
    • Config and profiles
    • JSON output and scripting
    • Devices
    • Overview
    • TypeScript
    • Python
    • Go
    • Elixir
    • Java
    • Rust
    • Secrets & network (all SDKs)
    • MCP overview
    • Install and connect
    • Tools reference
    • Sandbox tools
    • Computer tools
    • Deployment tools
    • Data tools
    • Platform tools
    • Advanced MCP setup
    • Tool server catalog

    • Authentication
      • Overview
      • Authentication
      • Requests, limits, and errors
      • Events (SSE)
      • Ownership and external attribution
      • Tenant configuration
    • Rate limits
    • Regions
    • Introduction
    • Interactive API explorer
    • Error codes
    • Events (SSE)
    • Region endpoints
    • Sandboxes overview
    • Create and run
    • Manage sandboxes
    • Previews
    • Exec
    • Streaming exec
    • Filesystem (computers)
    • Files (upload and download)
    • Files (tree, batch, watch)
    • Processes
    • Sandbox variables
    • Snapshots
    • Sandbox templates
    • Sandbox shares
    • Preview tokens
    • Computers
    • Desktop
    • OpenComputers
    • Cloud and BYOC
    • Services (systemd)
    • Agent runs
    • Agent run groups
    • Agent runtime profiles
    • Agent runtime variables
    • Deployments
    • Docker Host
    • Versions
    • Releases
    • Databases
    • Custom domains
    • Projects
    • Invites
    • Credits
    • Quotas
    • Usage rollup
    • Audit log
    • Webhooks
    • Organization branding
    • Organization preview domain
    • Organization events (SSE)
    • Network policy (per computer)
    • Egress vault and allowlist

  • Workspaces
  • Members and invites
  • API keys
    • Pricing and limits
    • Plans and limits
    • Billing and usage

  • Security overview
  • API key scopes and rotation
  • Audit log
  • Data retention and deletion
  • Shared responsibility
  • Security troubleshooting

  • Infrastructure options
    • Bring your own cloud
    • AWS
    • Google Cloud
  • External computer providers
  • Compare compute providers

  • MIOSA vs other platforms
    • Railway MIOSA vs Railway
    • Render MIOSA vs Render
    • Fly.io MIOSA vs Fly.io
    • E2B MIOSA vs E2B
    • Daytona MIOSA vs Daytona
    • Modal MIOSA vs Modal
    • Vercel MIOSA vs Vercel Sandbox
    • MIOSA vs Cloudflare Sandboxes
    • Freestyle MIOSA vs Freestyle
    • Runloop MIOSA vs Runloop
    • Blaxel MIOSA vs Blaxel
    • Beam MIOSA vs Beam
    • Boat MIOSA vs boat
    • CodeSandbox MIOSA vs CodeSandbox

  • Release notes
    • Move from Railway
    • Moving from Vercel
    • Migrating from miosa-sandbox
    • Availability and roadmap
    • Known limitations
    • Recovery playbooks
    • Evaluations
    • Benchmarks

    • Overview
    • Getting started
    • Sandboxes and computers
    • Browser and computer use
    • Agents
    • SOMA and runtimes (early access)
    • White-label and platform builders
    • Business OS and internal use
    • Security and compliance
    • Pricing and limits
    • Support
  • Troubleshooting
  • Support

FAQ

Short answers, grouped by topic, with a link to the full page for each.

Getting started

Where to start, how to authenticate, install the CLI, get an API key, attach IDs to a sandbox, and pick a region.

Sandboxes and computers

Sandbox vs Computer, the default sandbox, run length, states, pausing, and snapshots.

Agents

Supported harnesses, creating and running an agent, OSA, model keys, and where agents run.

Browser and computer use

Where browser automation runs, what starting a browser returns, CDP, live takeover, and billing.

Pricing and limits

What you are billed for, paused cost, named sizes, hard caps, rate limits, and credits.

Security and compliance

SOC 2, HIPAA, where data lives, how secrets are stored, retention, and publishable keys.

White-label and platform builders

White-label architecture, branded URLs, end-user auth, billing your customers, and branding.

Business OS and internal use

Running your own company on MIOSA, connectors, device choice, and platform checklists.

SOMA and runtimes

What SOMA is, the default runtime, how to select one, access requests, and availability.

Support

How to get help, what to include in a request, status and changes, and the community.

Was this page helpful?
Previous ←Benchmarks Next Getting started→

On this page


Was this page helpful?
MIOSA MIOSA

Documentation for building, running and shipping on MIOSA.

Subscribe to updates

Get started

  • Introduction
  • Quickstart
  • Install the CLI or SDK
  • Sandboxes or Computers?
  • First production app
  • What you can build
  • Compare

Products

  • Sandboxes
  • Computers
  • Agents
  • Docker Host
  • Databases
  • Storage
  • Domains

Developers

  • SDKs overview
  • TypeScript
  • Python
  • Go
  • Rust
  • Elixir
  • Java
  • CLI
  • MCP
  • API reference

Platform

  • Workspaces
  • Members
  • White-label
  • Billing
  • Security
  • Benchmarks

Resources

  • Changelog
  • Docs for LLMs
  • Full context
  • Status
  • Pricing
  • GitHub
  • Discord
  • LinkedIn
  • Instagram
  • Contact support

Legal

  • Trust Center
  • Terms
  • Privacy
  • Acceptable use
  • DPA
  • Subprocessors
  • Cookies
  • Security
  • SLA
  • Acknowledgements
© 2026 MIOSA
PrivacyTermsTrust CenterStatus
Checking system status

On this page