Getting Started with the Archive API
The Archive API gives you programmatic access to your workspace data — creators, content, engagement metrics, social profiles, and more. If you’re new to Archive, check out the Getting Started with Archive guide first.
All requests are made via a single GraphQL endpoint:
POST https://app.archive.com/api/v2Access & Limits
| Token provisioning | Self-serve. Generate your token directly from the Archive UI under Settings → API. Available on all 2026 plans that include API access. See Generating an API token for the step-by-step. |
| Token scope | Tokens are scoped at the workspace (or agency) level, not per user. For agency plans, a single token covers every workspace in the agency. |
| Sandbox | No sandbox environment is available — all queries run against your live workspace data |
| Rate limit | A credit budget sized by your plan, plus a flat ceiling of 5 requests per second — whichever you reach first. Credits measure backend compute (1 credit ≈ 1 ms), so a cheap lookup costs 5 and a filtered creator search costs ~350. Both ceilings are per workspace and never shared, including on agency plans. Over-budget requests return HTTP 429 with Retry-After and code RATE_LIMIT_EXCEEDED. Every response reports your remaining budget in the ratelimit header. Full breakdown per plan: Rate Limiting. |
| Time and timezones | All date/datetime inputs must be UTC, formatted as ISO 8601 with a trailing Z (e.g. "2026-03-01T00:00:00Z"). The API does not infer a timezone from the caller and does not accept offset notation. If your data lives in a local timezone, convert to UTC before sending. All datetime fields in the response are returned in UTC — convert to the user’s local timezone client-side when displaying. |
Choose how you want to use the API
There are two ways to interact with the Archive API. Choose the one that fits your workflow — or use both.
Using the MCP server
The Archive MCP server lets compatible AI assistants call the Archive API natively — no GraphQL to write, no .env file to wire up. Once connected, your assistant can search content, pull engagement, look up creators, manage views and collections, and more, all in plain language. It’s the fastest way to explore your workspace conversationally; for scripted automations, use the API client path below.
Server URL: https://app.archive.com/api/v2/mcp
Most agents connect with one-click browser sign-in — no API token to copy. Start from Settings → API → Connect an agent in Archive, pick your agent, and approve access in the browser.
| Creating a Custom Connector | Claude and ChatGPT need this one-time step, since Archive isn’t in either connector directory yet. |
| Connecting an Agent | Per-client sign-in — Claude, Claude Code, ChatGPT, Cursor, VS Code, and the config-file clients. |
| MCP Server | What the server can do once connected: all 53 tools, grouped by area. |
Using an API client
You can use any API client that supports GraphQL — such as Postman, Insomnia, or Hoppscotch. The steps below use Postman as an example, but the process is similar in any tool.
Authentication: The token goes in the Authorization header of every request:
Authorization: Bearer your_token_hereSetup:
- Create a new POST request to
https://app.archive.com/api/v2 - In the Authorization tab, set Auth Type to Bearer Token and paste your token
- In the Body tab, select GraphQL and paste your query
Next step: Workspaces — discover your workspaces and set up your workspace ID.