⚡

STT.ai i roto i te Desktop Claude, te Kaiārahi, me ētahi atu māngai MCP

Ka whakamātauria te oro/whitiāhua, te rārangi me te kōrero ki ngā whakamātautau o mua, te whakarāpopoto, te tātari, me te whakaputa ihirangi — i roto i te utauta AI e whakamahia ana e koe. Ko te pūnaha o te pūnaha o te tautuhi mō te STT.ai.

pip e whakatū ana i te sttai-mcp GitHub

✦ Ko te pūnaha pūnaha STT MCP anake e whakaae ana ki te whakamātautau me te pātai a Claude i roto i tētahi kōrero

Ko ētahi atu MCP whakamāoritanga (Otter, Fireflies, Granola, Read AI) e ako ana anake i runga i ngā tāruatanga i rokirokitia. Ko te pūnaha MCP o te STT.ai e tāpiri ana i te transcribe_url me te kōrerorero_me_whakaaro - whakapiri i tētahi pātahitanga YouTube ki Claude me te pātai i ngā pātai i runga i te mea, katoa me te kore e mahue te kōrero.

10 ngā utauta, e whakaritea ana hei whakamahi

transcribe_url

Whakataki + whakarerekē i tētahi URL — YouTube, pūngao podcast, faila hāngai.

chat_with_transcript

RAG Q&A i runga i tētahi tāruatanga me ngā whakahua pūtake me ngā tātaitai wā.

list_transcripts

Ko te rārangi pāpāho o ōna tāruatanga o mua, he tātari e ai ki te tūnga.

get_transcript

Ka whiwhi i tētahi tāruatanga me ngā wāhanga, ngā kaikōrero, me ngā metadata.

summarize_transcript

AI whakarāpopoto - i whakawāteatia, i whakahōu anō rānei.

analyze_transcript

Te āhua, ngā kaupapa, ngā mea, ngā mea mahi, ngā pātai, PII.

generate_from_transcript

Ko ngā tātai pukapuka, ngā pāpori, ngā tohutohu akoranga, ngā kāri, ngā tuhipoka whakaaturanga podcast.

export_transcript

Ka whakawātea hei txt, srt, vtt, json, csv rānei.

list_models

He tauira whakamāoritanga e wātea ana me ngā tohutoro WER.

list_languages

Ko ngā reo STT.ai e tautoko ana — 99 me te tatau.

Ka whakaritea i roto i te 2 min

1. Ka whakatū te pakiaka

pip install sttai-mcp

2. Ki te whiwhi i tōtou kī API

Ka tangohia tētahi mai i ōna Te pou ā-kōrero Ka whakawhiwhia rānei ki te wātea mo te wātea.

3a. Papamahi Claude

Ka whakakē i tōna pūāhua whakaritenga (~/Library/Application Support/Claude/claude_desktop_config.json rānei %APPDATA%\Claude\claude_desktop_config.json):

{
  "mcpServers": {
    "stt-ai": {
      "command": "sttai-mcp",
      "env": { "STT_API_KEY": "your-api-key-here" }
    }
  }
}

Ka tīmata anō te tauwāhi tauwāhi Claude. Me puta iho te utauta 10 STT.ai.

3b. Kāwai

Whakatauira → MCP → Tāpiri i tētahi pūnaha MCP hou, tāpiri i te JSON ōrite.

Ko ngā whakahau tauira

I roto i te Desktop a Claude, i muri i te whakaritenga:

"Transcribe https://www.youtube.com/watch?v=abc123 and summarize the key points."

Ko Claude e kī ana i te transcribe_url, kātahi ka whakarāpopoto i te hua.

"What are my last 5 transcripts about?"

E kī ana a Claude i te rārangi_transcripts (te tepe = 5) ā, ka pānuitia ia taitara.

"In transcript abc123, what did the speaker say about pricing?"

Ko Claude e kī ana i te kōrero me te whakamāoritanga, ā, ka whiwhi whakautu me ngā whakahua pūtake i te wā.

Matapihi pūtake (whakahaere MIT). He take, PRs, whakaaro e whakaaetia ana i GitHub.

github.com/sttaigit/sttai-mcp pypi.org/project/sttai-mcp