OpenAI Workspace Agents
Workspace Agent runs now use token-based pricing where pricing is active.
OpenAI Workspace Agents let ChatGPT Business, Enterprise, and Edu teams build agentic workflows for documents, spreadsheets, presentations, and business tasks, with token-based usage now tied to the Codex agentic usage pool.
Opens official website
Token-based Workspace Agent pricing
OpenAI's release notes state that as of July 6, 2026, Workspace Agent runs use token-based pricing where pricing is active. Usage is tied to input, cached input, and output tokens, and Business plans include Workspace Agents, ChatGPT for Excel, and ChatGPT for PowerPoint through the general Codex agentic usage pool as pricing takes effect.
- Billing model
- Token-based credits
- Usage pool
- General Codex agentic usage pool
- Admin focus
- Monitor agent runs, spreadsheet tasks, presentation tasks, and flexible credits
About OpenAI Workspace Agents
OpenAI Workspace Agents bring agentic workflows into ChatGPT Business, Enterprise, Edu, and related workspaces. Teams can create agents for recurring business tasks, spreadsheets, presentations, reports, and cross-functional workflows. The July 2026 buying change is pricing: OpenAI's Business and Enterprise/Edu release notes say Workspace Agent runs now use token-based credit pricing where pricing is active, with usage based on input tokens, cached input tokens, and output tokens. Business plans include Workspace Agents, ChatGPT for Excel, and ChatGPT for PowerPoint through the general Codex agentic usage pool as pricing takes effect for each feature. That makes Workspace Agents less like a flat chatbot feature and more like an agentic usage budget that admins need to monitor.
✅ Pros
- +Turns ChatGPT from a chat assistant into reusable workspace agents for teams
- +Connects agentic work to spreadsheets, presentations, and business workflows
- +Token-based credits make heavy agent usage easier to budget than fixed per-run estimates
- +Fits companies already standardized on ChatGPT Business, Enterprise, or Edu
❌ Cons
- −Agent runs can consume usage faster than ordinary chat workflows
- −Admins need to understand the Codex agentic usage pool and flexible pricing
- −Pricing and availability vary by plan and feature rollout timing
- −Not a replacement for governance, permissions review, or human approval on sensitive workflows
Key Features
Workspace Agent Runs
Create agents for recurring knowledge-work tasks, then track usage as token-based credits rather than treating every agent run as a flat-cost action.
Codex Agentic Usage Pool
Workspace Agents, Codex, and supported Office-style workflows share an agentic usage model, making cost visibility and admin controls more important for teams.
Spreadsheet and Presentation Workflows
ChatGPT for Excel/Sheets and ChatGPT for PowerPoint extend agentic usage into everyday business documents, so finance, ops, and marketing teams need clear review practices.
Pricing Plans
- ✓Workspace Agent runs
- ✓ChatGPT for Excel/Sheets where available
- ✓ChatGPT for PowerPoint where available
- ✓Usage counted through the Codex agentic usage pool
🏆 Our Verdict
OpenAI Workspace Agents are compelling for teams already using ChatGPT Business or Enterprise, but the July 2026 pricing change makes cost governance central. Treat agents like metered automation: define allowed tasks, monitor token use, set review paths for sensitive outputs, and budget the Codex agentic usage pool before broad rollout.
✅ Best For
Quick Info
- Pricing
- paid
- Category
- productivity
- Rating
- 4.4 / 5.0
- Last Reviewed
- June 2026