Sevana voice quality analysis
MCP server by Sevana · mcp
Sevana voice quality analysis. 1. Get a file onto the server: create_upload (returns a one-time URL; PUT the file there, e.g. with curl -T) or upload_inline (base64, small files). Use kind "wav" for recordings and "pcap" for packet captures. 2. Analyze it with its upload_id: - pvqa_analyze: MOS and impairments of one WAV recording, no reference needed. - aqua_compare: compare a degraded WAV with its reference WAV (MOS, percent, fault analysis). - pcap_analyze: VoIP calls in a .pcap/.pcapng: SIP calls, RTP streams, MOS per stream, loss, jitter, call health. Runs as a job; read the result with pcap_summary, pcap_get_call, pcap_get_stream, pcap_sip_messages, pcap_dtmf, pcap_network_stats. Reports never contain decoded audio. 3. Read sevana://docs/* for what the report fields mean. Without an API key the free tier allows 10 analyses per hour. Uploads and results are deleted after an hour.
Endpoint: https://mcp.sevana.biz/mcp
Verification: observed. Source: crawl.
Capabilities (22)
- aqua_compare
- create_upload
- delete_upload
- job_cancel
- job_status
- list_jobs
- pcap_analyze
- pcap_dtmf
- pcap_full_report
- pcap_get_call
- pcap_get_stream
- pcap_network_stats
- pcap_sip_messages
- pcap_summary
- pvqa_analyze
- pvqa_detector
- pvqa_timeline
- server_info
- upload_inline