Fluencify in ChatGPT and Claude.
Ask your assistant what UGC creators charge, which brands are taking creators right now, or for a brief your creators can film from. It answers with Fluencify's own data and links to the page that goes deeper.
Connect it
One address for every assistant. No account, no sign-in, and it only reads.
MCP server https://fluencify.io/mcp
Claude
Open Customize, then Connectors, and add a custom connector with the address above. Choose no sign-in, then switch Fluencify on in a chat from + > Connectors.
ChatGPT
Where your workspace allows custom MCP connectors (developer mode), add the address above as a connector.
Claude Code, Cursor, VS Code and other MCP clients
Add a remote MCP server with the address above, over Streamable HTTP. In Claude Code:
claude mcp add --transport http fluencify https://fluencify.io/mcp
What it can do
UGC creator rates
What UGC creators ask per video, content only, from public listings on a large freelance marketplace: the middle half of asking prices and the median, by niche and seller country, with the number of listings and sellers and the week beside it. A pick with too few listings gets the next range up, said as such. It never answers with Fluencify's own pay data.
Try asking “What do beauty UGC creators in the US charge per video?”
get_ugc_rate_rangesOpen creator campaigns
Brands taking creators right now, by niche and platform: the brand, its niche, where it posts and its open campaigns, each linked to its page here. Pay, countries and the brief are shown after sign-up.
Try asking “Which brands are paying creators for Instagram Reels right now?”
search_open_campaignsCreator brief generator
A brief creators can film from: deliverables, key messages, hooks to try, shot ideas, do's and don'ts, and the platform's ad-disclosure rule. It leaves pay and deadlines to you and never promises results.
Try asking “Write a UGC brief for my smart water bottle, for TikTok, to drive online sales.”
generate_creator_brief
What it reads and what it records
The same public data a visitor sees on this site, and a count of how the tools are used. More in our privacy policy.
- The tools read only what Fluencify already publishes: the brand-level list on the campaigns page. Nothing reads or returns anyone's account, deal terms, pay or contact details.
- For each call we count the tool, whether it worked, the platform, niche or goal you picked, and which assistant called. We do not record the text you typed or who you are.
- Briefs are written from the details you give, by a language model on Cloudflare Workers AI or from our template, and those details are used for nothing else. Identical requests may be answered from a cache.
- Calls are rate limited per network, so the server stays free for everyone.
For developers
MCP over Streamable HTTP with JSON responses, stateless. It speaks protocol 2026-07-28 (per-request metadata and server/discover) and the earlier revisions back to 2024-11-05 (the initialize handshake). Every tool is read-only and declares an output schema.
Discovery: the server card, the API catalog and the AI catalog. Questions or problems go to support.
List the tools from a terminal:
curl -s https://fluencify.io/mcp \
-H 'content-type: application/json' \
-H 'accept: application/json, text/event-stream' \
-d '{"jsonrpc":"2.0","id":1,"method":"tools/list"}'