Capy
Also known as: Capy AI, Scrapybara, capy.ai, Capy Corporation
Agent platform for engineering teams that runs parallel coding agents on any model in isolated cloud VMs or on the team's own machines, from issue to merged pull request, with Captains that orchestrate hundreds of agents and Slack, Linear and GitHub integrations. Built by Scrapybara, Inc., which moved from hosted computers for agents to Capy.
Capy is an agent platform for engineering teams from Scrapybara, Inc., the company that built Scrapybara, and it lives at capy.ai. It runs swarms of coding agents on any model, across Capy's cloud and the team's own machines. Each thread runs on its own Firecracker microVM, delegates work to subagents that nest up to three levels and can each use a different model, and keeps a plan the person can follow step by step. Captains are long lived orchestrator agents that can manage hundreds of agents, splitting a migration across subagents that each ship their own pull request.
Capy takes Linear issues and streams progress back into them, works in Slack threads, opens and reviews GitHub pull requests, and connects to Vercel, Tailscale and Revyl. It connects MCP servers, and through Tailscale it can reach private databases and internal services. Capy never merges unless asked in the thread, a built in Review agent reads each pull request in a real checkout and Autofix handles CI failures, review comments and merge conflicts, and a plugin marketplace, a skills library and a documented REST API let teams extend it and drive it from their own code.
A person picks the model for each thread from OpenAI, Anthropic, Google, xAI, Moonshot, Z.ai, DeepSeek, Alibaba, MiniMax and Meta models, can run one task on several models to compare approaches, and can connect existing Codex, Copilot or SuperGrok subscriptions or bring API keys and gateways. Agents run in Capy's cloud VMs, hosted in US East, on the customer's own Mac, Windows or Linux computer through the desktop app, or on any server through the Dock CLI, and threads can move between them mid task. Enterprise customers can run it in their own cloud on request.
Each machine has a desktop with Chrome, so agents test the apps they build in a real browser and record annotated videos, and from Pro up they test mobile builds on cloud iPhones and Android phones. Capy holds SOC 2 Type II, every plan includes an audit log, and single sign on is a $100 a month add on or included on Enterprise. Plans run from a free tier on the customer's own computers through Lite at $20 a month to Max at $200, with usage shared across the organization and no per seat fee.
Scrapybara built computers for agents, selling hosted Ubuntu, Windows and browser instances that computer use agents could control, with the Act SDK and Python and TypeScript SDKs. The company now builds Capy, and scrapybara.com is a one line page pointing to capy.ai.
From our cloud coding agents video
Devin, Factory, Jules, OpenHands: What Each Cloud Coding Agent Does Best
Our video on the 11 cloud coding agents in the index names what each one does best. For Capy, that is choice of model.
The player starts at Capy's part, 2:41 in. Recorded in October 2026; the profile here is live and updates as research is verified.
Vendor details
Canonical URL
https://capy.ai/
Category
Coding agent
Funding status
Independent. Built by Scrapybara, Inc., San Francisco, a Y Combinator company; the site footer reads Capy Corporation. Capy's first public release was the Capy Beta of 11 August 2026 and the changelog reached version 0.6.0 on 27 October 2026.
Company status
independent
Use cases & customers
Target customers
Deployment options
Integrations
GitHub App for pull requests and reviews, Slack with @Capy in threads, Linear delegation that creates a native agent session, Sentry report import, Tailscale for private networks, Vercel previews, Paper, OIDC federation to GCP, AWS, Azure and Vault, and MCP servers over HTTP or stdio with OAuth. Plugins from a marketplace bundle MCP servers and skills, and a REST API with an OpenAPI 3.1 spec drives projects, threads, automations, reviews and usage.
In practice
Your team tracks work in Linear. Capy takes an issue, plans it with its Captain agent, has Build agents carry it out on their own branch and VM, and opens a GitHub pull request that a person reviews before it merges.
You want several fixes moving at once. Capy runs parallel agents, each in its own isolated cloud VM, and a thread moves to needs attention when an agent is blocked on a decision, access or credentials.
Your security team needs a known control set. Capy has a SOC 2 Type II report, SSO with SAML and audit logging on the Enterprise plan, and enterprises can host it on premises or in their own cloud.
Sources & related URLs
Related / legacy domains
Research sources
Agentic Index coverage score
12.5 / 14 capabilities · 89%
| Integrations & Tool Calling | Full |
|---|---|
|
Capy acts inside the tools a team uses. A GitHub App opens and reviews pull requests, Slack threads take @Capy, delegating a Linear issue creates a native agent session that streams progress back, and Sentry reports can be imported. MCP servers connect over HTTP or stdio with OAuth, Tailscale ephemeral nodes reach private databases and internal services, Vercel previews, Revyl and Paper are integrated, and OIDC federation gives machines short lived GCP, AWS, Azure or Vault credentials. GitHub Enterprise Server is not supported. Sourcedocs.capy.ai/integrations/mcpread 2026-10-28 |
|
| Workflow Orchestration | Full |
|
A thread delegates work to subagents that nest up to three levels, on a shared or a fresh machine, each with its own model, and work such as a migration splits across subagents that each ship their own pull request. Captains are long lived orchestrator agents that direct crews of threads, keep their own Drive and can manage hundreds of agents, so a team can run thousands of agents at once, each task on its own branch and machine. Sourcedocs.capy.ai/subagentsread 2026-11-19 |
|
| Knowledge Grounding & RAG | Partial |
|
Agents work over a real checkout of the customer's repository inside their machine, nested AGENTS.md files load when a folder is first touched, a Review context page holds organization rules and files, and imported context such as Sentry reports and attached files lands on the agent's machine. No persistent codebase index, embeddings or retrieval layer is described, so what an agent knows of the code comes from reading the checkout during the task. Sourcedocs.capy.ai/instructionsread 2026-10-28 |
|
| Human Oversight & Guardrails | Full |
|
Capy never merges or enables auto merge unless asked in the thread, the Merge menu offers Enable auto merge and Force merge, and a root AGENTS.md can grant standing merge permission. The agent waits for a go ahead when told to, review mode runs Off, Once or on every push, paired devices ask before use, any agent can be stopped mid run, and a thread or Captain tags the user under Needs you when blocked on a decision, access or credentials. Sourcedocs.capy.ai/admin/securityread 2026-10-28 |
|
| Security, Identity & Governance | Full |
|
Capy holds SOC 2 Type II on every plan, audited and pen tested every year, with the report available through its trust center. Organizations carry Owner, Admin and Member roles, and every plan includes an audit log. Single sign on through SAML and OIDC with Okta, Google Workspace and Microsoft Entra ID, with required MFA, is a $100 a month add on for Lite, Pro and Max and included on Enterprise, and the desktop app signs in with SSO. Zero data retention is available on request, code is not used for training, and data is encrypted in transit and at rest. No SCIM is documented. Sourcecapy.ai/pricingread 2026-11-19 |
|
| Observability & Auditability | Full |
|
Each thread keeps its full transcript and tool output, shows the agent's plan, terminal output, file changes and diffs, and a status (active, waiting, needs attention, ready for review), with the machine's desktop streamed live into the thread pane. An Activity page and usage reports break spend down by member, model and thread, the API streams events, end to end test runs record annotated video, and every plan includes an audit log. Sourcedocs.capy.ai/async-workread 2026-11-19 |
|
| Memory & State Persistence | Partial |
|
A thread is a durable conversation stored on Capy's servers that holds the full history of its messages, machines, subagents and pull requests, survives archiving and resumes when woken. An opt in Memory experiment keeps what a user teaches Capy across threads, and review findings list the memory they rest on, but no scope, lifetime or way to delete entries is described. Drive holds AGENTS.md instructions at personal or organization level, which are instructions, not memory. Sourcecapy.ai/changelogread 2026-10-28 |
|
| Deployment & Data Residency | Full |
|
Each cloud thread runs on its own Firecracker microVM on Ubuntu 24.04, sleeping after about two minutes and deleted after 30 days asleep. The hosted service runs in one region, US East (New York), with no region choice. Threads can also run on the customer's own Mac, Windows or Linux computer through the desktop app or on any server through the Dock CLI, moving between machines mid task, and Enterprise offers running in the customer's own cloud on request, with OIDC federation issuing short lived cloud credentials to machines. Sourcedocs.capy.ai/machinesread 2026-11-19 |
|
| Prebuilt Agents, Templates & Packs | Full |
|
A plugin marketplace offers searchable, categorized plugins that bundle MCP servers and skills, with a Verified by Capy mark on plugins Capy has picked and tested, and named plugins include Firecrawl, ClickStack and Supermemory. A team installs plugins across the organization or per project or adds one from a GitHub repository. An Explore page lists plugins, skills and themes, and built in skills ship with the product beside the Review agent and Captains. Sourcedocs.capy.ai/integrations/pluginsread 2026-10-28 |
|
| Triggers & Channel Coverage | Full |
|
Automations carry one to twenty triggers: cron with a five minute floor, GitHub events, Slack message bursts, an incoming webhook and on demand runs from the API, and the API adds Linear triggers (issue created, status changed, end of cycle) and Sentry triggers. Delegating a Linear issue or mentioning @capy on a pull request comment starts work, and a thread wakes on its own when CI finishes, a review lands or a pull request merges. Sourcedocs.capy.ai/automationsread 2026-10-28 |
|
| Model Flexibility & Routing | Full |
|
A person picks the model per thread and per subagent from Anthropic, OpenAI, xAI, Google, Moonshot, Z.ai, DeepSeek, Alibaba, MiniMax, Meta and other makers, and can run one task on several models to compare. Teams can bring Anthropic, OpenAI, Google and xAI keys or gateways such as Vercel, OpenRouter and Cloudflare, and existing Codex, Copilot and SuperGrok subscriptions connect and draw on the provider's limits. Model tokens bill at each provider's list price with no markup. Sourcedocs.capy.ai/models-and-pricingread 2026-10-28 |
|
| APIs, SDKs & MCP Extensibility | Full |
|
A public REST API with bearer keys and an OpenAPI specification covers projects, threads, messages, tasks, automations, reviews and usage, so developers can start, message, interrupt and archive agent work from their own code and stream its events. MCP servers and plugins extend what the agents can reach, and the Dock CLI attaches any server as a machine for Capy threads. Sourcedocs.capy.ai/api-reference/overviewread 2026-10-25 |
|
| Testing, Debugging & Optimization | Partial |
|
A built in Review agent checks every pull request from a real checkout, labels each finding confirmed or investigate and comments on specific lines, and Autofix answers CI failures and review comments. Agents run the project's own tests and end to end checks in their machines, which tests the customer's code. On the agent side, the same task can run on different models to compare approaches, and there is no evaluation suite or regression testing of agent behavior. Sourcedocs.capy.ai/reviewread 2026-10-28 |
|
| Browser & Computer Use | Full |
|
Each cloud machine runs an XFCE desktop with Google Chrome that streams live into the thread pane, and the agents use native computer use and browser use for end to end testing, recording annotated videos of each run. From the Pro plan, agents also test iOS and Android builds on cloud iPhones and Android phones through Revyl. Capy does not describe approvals for purchases or form submissions made in the browser. Sourcedocs.capy.ai/machinesread 2026-10-28 |
|
The Agentic Index coverage score grades every vendor Full, Partial or Not documented against the same 14 buyer facing capabilities, from public evidence only. Each capability links to how all vendors in the index score on it. How this evidence is graded
Product updates
Version 0.8.0 of Capy adds an Identity and access page in organization settings where admins verify a domain by DNS record and connect Google Workspace sign in, sold as an add on, with Microsoft Entra ID, Okta and other SAML or OIDC providers by contact. Admins can turn on Require MFA with a deadline up to 30 days out, or a passkeys only mode, and every member can add passkeys, an authenticator app and backup codes and sign in to the desktop app with a passkey. Users can also pin notes on images and video frames in a thread and send them to the agent, and pull requests opened from Capy threads now arrive with their Guide already written.
Bears on: Security / enterprise
View sourceThe new Dock CLI in Capy 0.7.0 runs Capy without the desktop app, so a build server, cloud VM or spare Mac can host agent threads with the tools and credentials already on it. The same release moves Captains and Issues out of experiments and turns them on for every organization. A Captain directs a crew of threads, and Issues tracks the work they pick up, with blocking links, sub issues, labels, due dates and linked pull requests.
Bears on: Workflow orchestration
View sourceCapy 0.4.7 lets a Captain, the agent that coordinates a crew of other threads, run on one of the user's paired devices, and the Captain can start crew threads on any paired device or in the cloud. A Captain can also go back to being a regular thread.
Bears on: Deployment / data residency
View sourcePricing
Free on your own computers; Lite $20 a month, or $16 billed annually; Pro $100; Max $200; Enterprise custom
One subscription per organization with a shared monthly credit balance and no seat fee. Model tokens and cloud machine time draw the credits down.
Included quota
Lite includes $20 of model usage, 30,000 machine minutes and 250 review credits a month. Pro includes $100 of model usage, 150,000 machine minutes and 1,250 review credits. Max includes $200 of model usage, 300,000 machine minutes and 2,500 review credits. Free includes 100 review credits. A machine minute is one minute on a 1 CPU, 4 GB machine, so larger machines use minutes faster, from 1 a minute for Small to 32 for Big Guy, with Large at 4. Included usage resets monthly and does not roll over.
What is public
Capy is sold as one subscription per organization, with unlimited members and no seat fee. Lite is $20 a month, or $16 billed annually. Pro is $100 a month, or $80 billed annually. Max is $200 a month, or $160 billed annually, and use beyond Max moves to Enterprise, which has custom pricing.
Billing mechanics
Model tokens bill from Capy credits at each provider's list price with no markup. Tokens on the customer's own ChatGPT, Copilot or SuperGrok plan or API keys are never charged by Capy. Past the included minutes, machines bill hourly by size while awake, plus a 2 minute wind down: Small is $0.10 an hour, Large $0.40 and Big Guy $3.20. Sleeping machines cost nothing.
Cost watchouts
Parallel agents and larger machines use machine minutes quickly, and included usage does not roll over. Machine time is billed past the included minutes even when model usage runs on the customer's own key.
Variable cost rationale
The bill grows with the number of agents running in parallel, how long they run, the machine sizes chosen and the models they call.
Additional watchouts
Cloud phones on Pro and up bill $5.40 an hour for an iPhone or iPad and $3.60 for Android while awake. Each code review bills its model tokens at list price plus a Small machine while it runs, and higher effort reviews cost more.
Overage / add-ons
Beyond the included credits, balance tops up one for one from $5, with auto reload.
Sales call required
Mixed (some tiers require a call)
Free / trial
Free runs agents on the customer's own computers with their own model subscriptions and keys, captains, integrations and unlimited members. Lite's first week costs $1 with a verified card, one trial per card and per organization.
Commercial notes
Capy is built by Scrapybara, Inc., a Y Combinator company. Every plan includes SOC 2 Type II, an audit log, unlimited members and API access, and Enterprise adds a custom quota and running in the customer's own cloud on request.
Key ambiguities
Capy does not state the cost of a typical task, which depends on the models chosen and how long its machines run.
Cancellation / refund
Plans bill monthly or annually, and annual billing saves 20%. Enterprise adds a custom credit quota and invoicing with net terms.
Support SLA / resale
Enterprise includes SSO, 24/7 support with an SLA, Slack Connect, invoicing with net terms and zero data retention on request. SSO and SAML are a $100 a month add on for Lite, Pro and Max.
Missing data
Enterprise pricing is not published.
Related vendors
- Cognition — Maker of Devin, an autonomous AI software engineer
- 10Web — Agentic website platform whose AI agents build, host and manage…
- AgentUI — Managed AI app builder for operations teams that generates and hosts…
- Aider — Open source, model agnostic terminal coding agent that edits your…
- Amp — Amp is a coding agent from Amp Frontier Corporation that runs each…
- Anthropic Claude Code — Claude Code is Anthropic's agentic coding tool
Alternatives to Capy
Other products that do the work Capy does, and what each one is.
- FactoryAgent native development platformCapy vs Factory →
- TemboCloud execution platform for coding agents you already useCapy vs Tembo →
- Google JulesCapy vs Google Jules →
- CursorAI native editor and agent platform
- Google Antigravity
- AmpCoding agent