Hover to scrollAsk a research question, get a cited brief back. Research Desk searches the live web, reads the pages it finds, and writes an answer in which every claim carries the index of a source it actually fetched — anything it could not attribute is dropped before delivery, and the count of what was dropped ships with the brief.
- Three depths that buy real work, not longer prose: quick 15 credits (1 search, ~6 sources), standard 35 (3 angles, ~14 sources read in full), deep 75 (6 angles, long reads, plus a second pass that searches the gaps the first pass left open)
- Every citation is re-checked against the pages actually retrieved; a claim citing nothing real is dropped rather than repaired, and unsupported briefs are refunded in full — never settled
- Narrow a question by recency window or by domain (prefix with '-' to exclude); each source comes back with a verbatim quote so checking a claim is a glance
- Built for AI agents day one: rsd_live_ REST keys, an MCP server (works with the Ounie AI Team via ?api_key=), keyless x402 pay-per-call in USDC on Base, and the fleet master key
For agents
Works with Claude, ChatGPT, and any MCP client. One server, every app: paste the gateway config and the agent can discover and run Research Desk — and the rest of the store — through ounie_list_apps → ounie_describe_app → ounie_invoke_app.
{
"mcpServers": {
"ounie": {
"url": "https://mcp.ounie.com"
}
}
}Scoped endpoint — just this app. Point a client here when the agent should only ever see Research Desk's tools:
{
"mcpServers": {
"research": {
"url": "https://research.ounie.com/api/mcp"
}
}
}To spend your Ounie credits on the scoped endpoint, append your app key to the URL as ?api_key=… — the header-less form clients like the Ounie AI Team need.
Credits used
| Research brief — quick | 15 credits per call |
| Research brief — standard | 35 credits per call |
| Research brief — deep | 75 credits per call |