Factory
Also known as: Factory Droid, Droids, Droid CLI, Droid Computers, Missions, Agent Readiness
Agent-native development platform whose Droids take a ticket from Linear or Jira and run it end to end, coordinated by Missions across parallel agents, from the app, CLI, headless CI or managed cloud sandboxes.
Factory is an agent-native software development platform built around the idea that engineering work should be delegated to autonomous agents rather than typed line by line. Founded in 2023, the company calls its agents Droids, and they are designed to take on the full software lifecycle: planning, writing code, testing, reviewing, documenting, and shipping, along with larger jobs like refactors, migrations, and incident response. Unlike an in-editor assistant that suggests the next line, a Droid runs in the background and executes a whole task end to end while the developer works on something else.
A Droid can delegate focused work to custom droids, subagents with their own prompt, model and tool policy, and Mission Mode coordinates several agents across a larger project. AutoWiki generates structured wikis of each repository's architecture, modules and conventions and refreshes them on every push, and automations start Droid on a schedule, from a Slack message, on a GitHub event or from a webhook.
The natural unit of work is often a ticket. Factory treats Linear and Jira as first-class entry points, pulling a ticket's description, acceptance criteria, comments, and linked context directly into the agent before it starts. Droid runs on developer machines through the desktop app and CLI, in CI through headless droid exec, and on managed cloud Droid Computers, and enterprises can deploy it in hybrid or fully airgapped patterns with their own model endpoints. Many droids can run in parallel, which is the point for large migrations and refactors that would otherwise be sequential.
Factory is model-agnostic: customers pick among models from Anthropic, OpenAI, Google, xAI and Factory's own pool, let a router choose per task, or bring their own keys and local models. The platform reaches developers through VS Code, JetBrains, the terminal, Slack, and pull-request automation, and is aimed at enterprise engineering teams, including regulated industries, that need autonomous, long-horizon work spanning many repositories and systems.
Vendor details
Canonical URL
https://factory.ai/
Category
Coding agent
Funding status
Independent. Founded 2023. $150M Series C led by Khosla Ventures at a $1.5B valuation in April 2026, with Sequoia, Blackstone and Insight participating, following a $50M Series B led by NEA in September 2025 with Sequoia, J.P. Morgan and Nvidia; around $220M raised. Named enterprise customers include Nvidia, Adobe, EY, Palo Alto Networks and Adyen.
Company status
independent
Use cases & customers
Target customers
Deployment options
Integrations
Linear and Jira are first class ticket entry points, pulling description, acceptance criteria, comments and linked context into the agent before work starts. GitHub and GitLab integration covers pull request automation and Actions, Slack triggers work conversationally, and Notion and Sentry are documented alongside VS Code and JetBrains IDE integrations. MCP connects arbitrary external tools and data sources via droid mcp add, hooks run shell commands around agent lifecycle events, and plugins bundle commands, droids, skills, hooks and MCP servers as distributable packages from an official marketplace at Factory-AI/factory-plugins, with Claude Code plugin layouts translated on install.
In practice
You have a sprawling migration that would take one engineer weeks of repetitive edits. Factory runs multiple Droids in parallel, each taking a slice of the work end to end, instead of grinding through it sequentially.
Your team lives in Linear and Jira. Factory treats a ticket as the unit of work, pulling its description, acceptance criteria, and linked context into a Droid that implements the change and opens a pull request.
You want autonomous agents but your code can't leave your environment. Factory's Droids run in sandboxed cloud workspaces or locally through a desktop app and CLI, fitting teams with or air-gapped requirements.
Sources & related URLs
Related / legacy domains
Agentic Index coverage score
12.5 / 14 capabilities · 89%
| Integrations & Tool Calling | Full |
|---|---|
|
MCP connects Droid to the customer's own tools and data, remote delegations take work from Slack, Linear and Jira, and automations act on GitHub events and inbound webhooks, with org policy deciding which models, tools and integrations Droid may use. Sourcedocs.factory.ai/harness/mcp, remote-delegations and enterprise/identity-and-accessread 2026-09-29 |
|
| Workflow Orchestration | Full |
|
Custom droids are subagents with their own system prompt, model and tool policy that Droid delegates focused tasks to, and Mission Mode coordinates multiple agents across a larger project beside Normal and Spec modes, with worktrees for parallel sessions. Sourcedocs.factory.ai/harness/subagents and autonomy-and-safety/specification-moderead 2026-09-29 |
|
| Knowledge Grounding & RAG | Full |
|
AutoWiki generates structured wikis of each repository's architecture, modules, APIs and conventions, keeps them current by regenerating on every push through a CI workflow, and makes them searchable in the Factory App and through the AutoWiki API, a maintained retrieval structure over the customer's code. The limit: the Software Factory dashboard that groups AutoWiki is in private preview, while /wiki runs from the Droid CLI. Sourcedocs.factory.ai/software-factory/wiki/overviewread 2026-09-29 |
|
| Human Oversight & Guardrails | Full |
|
Autonomy levels (Off, Low, Medium, High) decide what Droid may do without confirmation, and permission rules allow, require approval for, or block matching shell commands, with an ask rule requiring approval even at High autonomy; org managed settings enforce these as hard policy, beside OS level sandboxing and secret scanning. Sourcedocs.factory.ai/autonomy-and-safety/permission-rules and auto-runread 2026-09-29 |
|
| Security, Identity & Governance | Full |
|
Factory lists SOC 2 Type II, ISO 27001 and ISO 42001 with reports, architecture documents and subprocessors in its Trust Center at trust.factory.ai, and documents SSO over SAML or OIDC, SCIM directory sync, three organization roles flowing from the identity provider, service accounts with their own API keys and an organization audit log. Sourcedocs.factory.ai/enterprise/compliance-audit-and-monitoring and enterprise/identity-and-accessread 2026-09-29 |
|
| Observability & Auditability | Full |
|
Droid exports OpenTelemetry to the customer's collector: each user turn produces a trace whose child spans are the model calls and tool runs, with droid.tool and droid.message spans when message content logging is on, plus tool, MCP and blocked command metrics tagged by session and user; an organization audit log records who did what. Sourcedocs.factory.ai/enterprise/telemetry/data-reference and compliance-audit-and-monitoringread 2026-09-29 |
|
| Memory & State Persistence | Partial |
|
Sessions are resumable across the app, CLI, web and mobile and drivable through the Sessions API, and AGENTS.md and hierarchical settings carry configuration across sessions. Session state lives within one session, AGENTS.md is instructions rather than memory, and no memory layer with a stated scope and lifetime is documented. Sourcedocs.factory.ai/api-reference/sessions and llms.txtread 2026-09-29 |
|
| Deployment & Data Residency | Full |
|
Three documented deployment patterns: cloud managed, with Droid on developer machines and CI and LLM traffic to the customer's own providers or gateways; hybrid, on the customer's VMs, containers and CI runners; and fully airgapped, with in network model endpoints and no Factory cloud at runtime; an EU pattern and Enterprise data residency, dedicated compute and on premises options are named. Sourcedocs.factory.ai/enterprise/network-and-deployment and pricing/organizationsread 2026-09-29 |
|
| Prebuilt Agents, Templates & Packs | Full |
|
Plugins package skills, commands, custom droids, hooks and MCP servers, and Factory maintains an official marketplace at Factory-AI/factory-plugins naming droid-control (terminal, browser and computer automation for QA and demos), droid-evolved (skills for navigation, writing, design and frontend work) and security-engineer (security review, threat modeling and vulnerability validation), each doing its own job; organizations can run internal approved marketplaces. Sourcedocs.factory.ai/harness/pluginsread 2026-09-29 |
|
| Triggers & Channel Coverage | Full |
|
Automations start a Droid session on a schedule (plain words or cron), a Slack message, a GitHub event or an inbound webhook, running on a Droid Computer, the desktop app or GitHub Actions, and Droid Exec publishes GitHub Actions workflows for PR review and weekly cron runs, so work reaches the agent with no person starting it. Sourcedocs.factory.ai/software-factory/automations and droid-exec/overviewread 2026-09-29 |
|
| Model Flexibility & Routing | Full |
|
The customer picks among Factory hosted models from Anthropic, OpenAI, Google, xAI and Factory's own pool, lets the Factory Router choose per task, or brings its own OpenAI, Anthropic, open source or local models by key, and enterprise controls restrict which models, base URLs and providers a fleet may use. Sourcedocs.factory.ai/models and model-independence/byokread 2026-09-29 |
|
| APIs, SDKs & MCP Extensibility | Full |
|
A REST API at api.factory.ai with bearer API keys creates and drives Droid sessions, wiki runs, CI automations and readiness reports (the Sessions API is enabled for selected organizations), beside the Droid SDK and headless droid exec for scripts and CI. Sourcedocs.factory.ai/api-reference/sessions and droid-exec/overviewread 2026-09-29 |
|
| Testing, Debugging & Optimization | Partial |
|
The Agent Readiness model scores how ready the customer's repository is for agents, code review, security review and QA automations test the customer's code, and Factory publishes its own benchmark results. None is a harness, scored test set or optimization loop for evaluating Droid's behavior on the customer's work. Sourcedocs.factory.ai/agent-readiness/overview and benchmarksread 2026-09-29 |
|
| Browser & Computer Use | Partial |
|
Droid Control, an official Factory plugin, lets Droids launch and operate terminal CLIs, web apps and Electron apps through end to end flows, clicking, typing and capturing screenshots for QA and demo evidence. It runs on runtime tools the customer installs, agent-browser among them, a third party browser engine wired in rather than one the platform runs. Sourcedocs.factory.ai/software-factory/droid-controlread 2026-09-29 |
|
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
Recent platform changes
Factory's September 19 release, CLI v0.223.0 and Desktop v0.180.0, changes the app's default session autonomy from Off to Medium when a user has not chosen a level. Medium permits file edits and reversible workspace operations such as dependency installation, builds, and local commits without repeated approval, subject to organization and safety controls. The same release adds an organization managed Databricks connector and requires confirmation for more commands that write to disks.
Bears on: Human approval / guardrails
View sourceFactory updated its Terms and Conditions for Individual Plans, detailing the licensing agreement for its AI coding platform and development tools. The revised terms explicitly state that customers grant Factory a non-exclusive license to access their connected codebases across platforms like GitHub, GitLab, and Bitbucket.
Bears on: Security / enterprise
View sourceFactory released Droid Shield 2.0 with two fine-tuned models for secret detection during autonomous commits, improving accuracy on real credentials versus false alarms while cutting latency and cost. Model weights were released for security research.
Bears on: Security / enterprise
View sourcePricing
Pro $20/mo · Plus $100/mo · Max $200/mo · Teams $60/mo + $40/seat · Business and Enterprise custom
hybrid
Included quota
Pro $20, Plus $100 (about five times Pro's usage) and Max $200 (about ten times): Standard Usage within 5 hour, 7 day and 30 day rolling rate limits, the free Droid Core pool and a free BYOK allowance. Teams: Pro rate limits per seat, up to 10 seats. Business and Enterprise: custom shared usage limits.
What is public
factory.ai/pricing and docs.factory.ai/pricing publish Individual plans Pro $20, Plus $100 and Max $200 a month, each with the Factory App, Droid CLI and Droid SDK and usage governed by 5 hour, 7 day and 30 day rolling rate limits, and Teams at $60 a month per team plus $40 a month per seat for up to 10 seats with Pro rate limits per seat; Business (up to 150 seats) and Enterprise (unlimited) are custom, adding SSO and SCIM, admin controls, ZDR, audit logging and, on Enterprise, data residency, dedicated compute and on premises options.
Billing mechanics
Sessions consume Standard Usage based on model usage (per model multipliers) plus Droid Computer compute, within three independent rolling rate limits; after Standard Usage, a free Droid Core pool of open weight models runs on its own limits; Extra Usage is prepaid credit from $10 that does not expire, and Missions require it; BYOK is free up to an allowance on Individual plans. Teams is billed month to month; Business and Enterprise use shared usage under contract terms.
Cost watchouts
Three independent rolling rate limits; Standard Usage does not roll over; Missions require Extra Usage; premium model multipliers drain usage faster than Droid Core; Droid Computer compute also consumes usage.
Variable cost rationale
A flat plan price with usage inside rolling rate limits, and optional prepaid Extra Usage beyond them, so spend is bounded unless Extra Usage is topped up; model choice and Droid Computer compute drive consumption.
Additional watchouts
Three rolling rate limits must all have headroom to send a request; Standard Usage does not roll over; Missions pause at a rate limit unless Extra Usage is on; premium models draw usage faster than Droid Core.
Overage / add-ons
When Standard Usage is exhausted, Droid Core models run on a separate free Rate-Limit pool; beyond that, prepaid Extra Usage credits ($10 min, roll over) are consumed. Standard Usage does not roll over month to month; purchased Extra Usage does.
Sales call required
Mixed (some tiers require a call)
Free / trial
No free tier; after Standard Usage runs out, a free Droid Core pool of open weight models keeps working on its own rate limits
Lowest paid plan
Pro $20/mo
Commercial notes
Individual and Teams plans are bought self serve; Business and Enterprise are sold through sales with custom limits and contract terms.
Key ambiguities
Standard Usage allotments per plan are not published and consumption varies with model multipliers and task size; Business and Enterprise are custom.
Cancellation / refund
Individual plans are self serve monthly; Teams is month to month; Extra Usage credits do not expire; Business and Enterprise are on contract terms. Refund terms are not stated.
Support SLA / resale
Business adds dedicated onboarding and support; Enterprise adds a dedicated account manager, customer engineer and priority SLAs (docs.factory.ai/pricing/organizations).
Missing data
The size of each plan's Standard Usage is not stated; Business and Enterprise pricing is custom.
Related vendors
- Cognition — Maker of Devin, an autonomous AI software engineer
- 10Web — Agentic website platform whose specialized AI agents build, host,…
- AgentUI — Managed AI app builder for operations teams that generates and hosts…
- Aider — Open source, model agnostic terminal coding agent that edits your…
- Anthropic Claude Code — Anthropic's agentic coding system across terminal, desktop, IDE, web…
- AppFactor — Agentic platform that runs persistent agents across an enterprise…
Alternatives to Factory
The closest documented capability profiles to Factory among coding agents tracked by Agentic Index, ordered by similarity on the same 14 point evidence the rankings use. No vendor pays for placement.
- Augment Code12.0 / 14A lighter documented profile than FactoryFactory vs Augment Code →
- Cognition13.0 / 14Fuller documented coverage on Browser & Computer UseFactory vs Cognition →
- Warp13.0 / 14Fuller documented coverage on Browser & Computer Use
- Zencoder13.0 / 14Fuller documented coverage on Testing, Debugging & Optimization
- Cursor12.5 / 14Fuller documented coverage on Browser & Computer UseFactory vs Cursor →
- GitHub Copilot13.5 / 14Fuller documented coverage on Memory & State Persistence and Browser & Computer UseFactory vs GitHub Copilot →
Similarity is computed from each vendor's Agentic Index coverage score evidence, axis by axis, not from the totals. How this evidence is graded