How much does Coworker cost?

Vendor blog content cites thirty dollars per user per month with the agent builder included; enterprise deployments are scoped with the implementation team. Confirm current terms with the vendor since figures appear in blog comparisons rather than a formal pricing page.

Back to vendors

Coworker

Q: Who is Coworker for?

Mid market and enterprise operations teams, Sales, customer success, and IT teams, CIO organizations standardizing on an agent platform

Also known as: Coworker.ai, OM1, OM2

Visit site

Enterprise operations agentindependentVerified 2026-07-08

Enterprise AI agent platform pairing an organizational memory knowledge graph (OM1/OM2) with model routing, fifty plus read and write connectors, a no code agent builder, and approval gates.

Coworker is an enterprise AI agent platform built on three pieces of infrastructure: a portable context layer, an intelligent model routing layer, and enterprise ready open source models. Its Organizational Memory (OM1, with OM2 adding MCP connectivity) continuously indexes connected tools into a permission aware knowledge graph of atomic facts and entity relationships, giving agents accurate company context across more than fifty read and write connectors spanning CRM, comms, support, docs, code, and data warehouses.

Agents run around the clock on schedules or event triggers, execute end to end work such as CRM updates, ticket creation, and report drafting, and wait for approval before acting where gates are configured. A no code Agent Builder defines what agents do, which tools they use, and when they run; the router scores each task across closed, open source, and self hosted models and sends it to the best one, with automatic adoption of new models. Governance includes permissions inherited from existing tools, approval workflows, credit limits, escalation logic, and full audit trails.

Founded by former Uber executives Alex Calder and Bradford Church in San Francisco, Coworker launched publicly in May 2025 with a thirteen million dollar seed led by Jeff Huber at Triatomic Capital. It is SOC 2 Type II audited, GDPR compliant, and CASA Tier 2 verified, with cloud, private cloud, on premises, and air gapped deployment plus VPC peering and bring your own model support.

Vendor details

Canonical URL

https://coworker.ai

Subcategory

Enterprise AI agent platform with organizational memory

Funding status

Seed; thirteen million dollars led by Jeff Huber (Triatomic Capital), May 2025 per company announcement

Company status

independent

Use cases & customers

Primary use cases

Autonomous agents for recurring operations workMeeting driven follow through (CRM updates, tickets, summaries)Grounded enterprise search and Q&ADocument, deck, and dashboard generationCustom agent building without code

Target customers

Mid market and enterprise operations teamsSales, customer success, and IT teamsCIO organizations standardizing on an agent platform

Deployment options

SaaSPrivate cloudOn-premAir-gappedVPC peering

Integrations

Fifty plus (marketed up to one hundred plus) OAuth connectors with read and write access: Slack, Salesforce, HubSpot, Jira, Gmail, Google Docs, GitHub, BigQuery, Snowflake, and more. Agents inherit existing permissions; OM2 exposes organizational memory to any MCP enabled client.

In practice

A revenue operations team deploys agents that join meetings, update Salesforce, create Jira tickets, and send follow ups automatically with approval gates on outbound actions

An IT service desk grounds responses in organizational memory, saving thousands of hours in the first deployment phase per a customer testimonial

An engineer connects OM2 over MCP so agents in other clients get company context with fewer tool calls

Sources & related URLs

Research sources

https://coworker.ai/ https://coworker.ai/platform https://coworker.ai/agents https://www.prnewswire.com/news-releases/coworkerai-launches-the-first-ai-agent-with-deep-company-context-backed-by-13m-seed-round-302459262.html

Capability coverage

12.5 / 14 capabilities · 89%

Integrations & Tool CallingFifty plus (marketed up to one hundred plus) read and write OAuth connectors across CRM, comms, support, docs, code, and data warehouses including Slack, Salesforce, Jira, Gmail, BigQuery, and Snowflake; agents inherit existing permissions. coworker.ai homepage and platform pages retrieved 2026-07-08.	Full
Workflow OrchestrationOrchestrates multiple agents from a single control plane with configuration across any channel or workflow, long running agents with triggers, and end to end multi step execution. coworker.ai platform and agents pages retrieved 2026-07-08.	Full
Knowledge Grounding & RAGOM1 builds a living knowledge graph by continuously indexing connected tools, decomposing information into atomic facts with permission aware recall so every agent answer is grounded, not generic. coworker.ai platform page retrieved 2026-07-08.	Full
Human Oversight & GuardrailsBuilt in permissions, approval workflows, escalation logic, credit limits, and full audit trails, with agents that wait for approval before acting where gates are configured. coworker.ai homepage and agents pages retrieved 2026-07-08.	Full
Security, Identity & GovernanceSOC 2 Type II audited, GDPR compliant, CASA Tier 2 verified, permission aware agents that respect existing access controls without privilege elevation, US hosted with no training on customer data. coworker.ai homepage and May 2025 PR retrieved 2026-07-08.	Full
Observability & AuditabilityEvery action logged and every decision auditable, with approvals, monitoring, evaluation, and optimization named as platform capabilities and self updating filterable dashboards on live data. coworker.ai homepage and platform pages retrieved 2026-07-08.	Full
Memory & State PersistenceOM1 gives agents a persistent knowledge graph to track progress, learn from corrections, and improve with entity aware, permission aware recall; OM2 adds shared skills and artifact templates over MCP. coworker.ai platform page retrieved 2026-07-08.	Full
Deployment & Data ResidencyCloud, private cloud, on premises, or air gapped deployment with VPC peering and bring your own model supported out of the box. coworker.ai platform page retrieved 2026-07-08.	Full
Prebuilt Agents, Templates & PacksShips a catalog of prebuilt agents (a churn prevention agent plus dozens more), role tailored agent suggestions on connect, and shared skills and artifact templates via OM2. coworker.ai homepage and platform pages retrieved 2026-07-08.	Full
Triggers & Channel CoverageAgents trigger on schedule (hourly, daily, weekly) or on real time events (new Slack message, Jira update, CRM change, calendar event), and run embedded in Slack, meetings, and boards. coworker.ai agents page retrieved 2026-07-08.	Full
Model Flexibility & RoutingIntelligent routing scores each task across closed, open source, and self hosted models (OpenAI, Anthropic, Google, Llama, Mistral, MiniMax) and sends it to the best one, with automatic adoption of new models and per task manual control. coworker.ai homepage and platform pages retrieved 2026-07-08.	Full
APIs, SDKs & MCP ExtensibilityOM2 organizational memory is available in Coworker's apps or any MCP enabled client, and a no code Agent Builder plus cloud coding sandbox extend the platform. coworker.ai homepage and platform pages retrieved 2026-07-08.	Full
Testing, Debugging & OptimizationThe control plane lists evaluation and optimization alongside approvals and monitoring, and Ambient Learning lets agents observe and replicate workflows; a dedicated test or simulation harness is not detailed. coworker.ai platform page retrieved 2026-07-08.	Partial
Browser & Computer UseAgents act through connectors and a coding sandbox; no browser or computer use capability is documented. coworker.ai pages retrieved 2026-07-08.	Unable to verify

Recent platform changes

No recent material changes tracked yet.

View all changes for Coworker →

Pricing

Thirty dollars per user per month (per vendor blog; agent builder included)

per user per month; enterprise scoped separately

Public — partialLow variable cost

What is public

A per user per month figure is repeated across vendor blog comparisons with the agent builder included; enterprise pricing is contact based.

Billing mechanics

Per user per month for the standard offering; enterprise deployments scoped with the implementation team without months long professional services per the vendor.

Cost watchouts

Enterprise deployments (private cloud, on premises, air gapped) are scoped separately and likely priced above the per seat figure; token or compute costs from model routing may factor into enterprise contracts.

Variable cost rationale

Per seat pricing keeps costs predictable, though model routing across providers and long running agents imply some underlying compute cost that a seat price may or may not fully absorb; enterprise scoping adds implementation variability.

Additional watchouts

Confirm the current per seat rate directly; enterprise and self hosted options will price differently.

Sales call required

Mixed (some tiers require a call)

Free / trial

Run your first agent in minutes on connect; no explicit free tier documented

Lowest paid plan

Thirty dollars per user per month per Coworker blog comparisons

Key ambiguities

The thirty dollar per user figure appears in Coworker's own blog listicles rather than a formal public pricing page, so it may be a representative rather than a committed rate.

Missing data

A canonical public pricing page; whether a free tier or trial exists; enterprise price points.

Verified 2026-07-08

Official pricing page

Data confidence: high