The reason to use Jasper rather than a raw model is that someone has already encoded the brand voice, the audiences and the style rules. Exposing those as separate readable tools means another agent can borrow that work — pulling the style guide into its own generation — without going through Jasper's generator at all. That is more useful than the generation tool for anyone with their own pipeline.
Jasper's endpoint, seven tools. Five read the brand context a workspace has configured; two act on it. Jasper frames the set as a subset of their API tailored for agents.
- get-jasper-brand-voices, get-jasper-audiences, get-jasper-style-guides — the workspace's configured brand context, read-only
- search-knowledge-base — the documents behind that context
- get-jasper-agents — the marketing agents available
- run-jasper-agent — execute one; this is the write and generative action
- generate-content — produce content using the configured voices, audiences, style rules and knowledge
OAuth 2.0 with PKCE and dynamic client registration, scope mcp, or a static key in an X-API-KEY header.
Paste a URL, then authorize — add the endpoint to your client, then approve the OAuth consent screen