Free tier, no sign-up
The Sevana MCP server lets AI agents and assistants run Sevana’s voice quality analyzers on their own. An agent that speaks the Model Context Protocol connects to one address, uploads a recording or a packet capture, and gets back the same analysis reports our products produce: MOS scores, impairments, call and stream statistics.
https://mcp.sevana.biz/mcp
MOS of a single WAV recording, with no reference needed. Also a timeline of quality over the call and the individual impairment detectors (echo, noise, clipping, dead air and more). About PVQA
Full-reference comparison of a degraded recording against the original, wideband: MOS, similarity percentage and the faults found. About AQuA
Analysis of a network capture: SIP calls, RTP streams with MOS, loss, jitter and delay, SIP messages and DTMF. The same engine as Pcaptix.
The server uses the Streamable HTTP transport, so any MCP client that supports remote servers can use it. Without a key it runs on the free tier.
claude mcp add --transport http sevana https://mcp.sevana.biz/mcp
{
"mcp": {
"sevana": { "type": "remote", "url": "https://mcp.sevana.biz/mcp" }
}
}
{
"mcpServers": {
"sevana": {
"url": "https://mcp.sevana.biz/mcp",
"headers": { "Authorization": "Bearer YOUR_API_KEY" }
}
}
}
Leave out the headers line to use the free tier. Then ask the agent in plain words, for example: “Estimate the voice quality of call.wav” or “Analyze capture.pcap and tell me what is wrong with the calls”. The server tells the agent how to upload the file.
| Free tier (no key) | With an API key | |
|---|---|---|
| Analyses | 10 per hour | by plan, up to 1000 per day |
| WAV files | up to 10 MB, 2 minutes | up to 100 MB, 30 minutes |
| PCAP files | up to 10 MB | up to 200 MB |
For an API key, higher limits or an on-premises installation, contact us.