API keys and MCP
Create API keys, call every public REST endpoint, and connect the official Speechdash MCP server.
Use Settings → API to create keys and connect assistants. The same sh_live_… key authorizes the public REST API and the official MCP server. Connectors under Settings → Connectors are a different feature: they attach third-party MCP servers to in-app chat.
Create a key
- Open Settings → API.
- Click New key, name it (for example
Production server), and copy it immediately. Only the prefix is stored; the full secret is shown once. - Send
Authorization: Bearer sh_live_…from your server. Never put a key in a browser, a mobile app, or a public repository.
You can keep up to 10 active keys. Revoke a leaked key from the same page.
Verify a key with GET /v1/me (email, plan, and wallet). Base URL: https://api.speechdash.com/v1.
REST endpoints
Every public HTTP route:
| Method | Path | Purpose |
|---|---|---|
GET | /health | Service health (no API key; not under /v1) |
GET | /v1/openapi.json | OpenAPI 3.1 contract |
GET | /v1/me | Account, plan, and wallet |
GET | /v1/documents | List library documents, including finished transcripts |
POST | /v1/documents | Create a text document (source=api) |
GET | /v1/documents/{document_id} | Read a document and its text |
PUT | /v1/documents/{document_id} | Update title, text, language, is_archived, or visibility |
DELETE | /v1/documents/{document_id} | Delete a document |
POST | /v1/documents/{document_id}/translate | New translated version (same credits as MP3 export; transcript clocks copied unchanged) |
GET | /v1/documents/{document_id}/export | Download pdf, docx, txt, csv, srt, or vtt |
POST | /v1/audio/speech | Synthesize up to 5,000 characters (JSON + audio) |
POST | /v1/audio/stream | Stream WAV, up to 20,000 characters |
POST | /v1/audio/stream/with-timestamps | SSE with per-sentence audio and marks |
GET | /v1/voices | Voice catalog |
GET | /v1/voices/{voice_id} | One voice preset (for example F1) |
The API does not start audio or video transcription jobs. Finished jobs appear as library documents (source=transcription). Interactive try-it-out pages: Account, Documents, Audio, Voices.
MCP tools
The official server at https://mcp.speechdash.com/mcp exposes one tool per library or speech action. ChatGPT uses OAuth. Cursor and Claude Desktop use Streamable HTTP with your personal API key.
| MCP tool | REST |
|---|---|
speechdash_get_account | GET /v1/me |
speechdash_list_documents | GET /v1/documents |
speechdash_get_document | GET /v1/documents/{document_id} |
speechdash_create_document | POST /v1/documents |
speechdash_update_document | PUT /v1/documents/{document_id} |
speechdash_translate_document | POST /v1/documents/{document_id}/translate |
speechdash_export_document | GET /v1/documents/{document_id}/export |
speechdash_delete_document | DELETE /v1/documents/{document_id} |
speechdash_list_voices | GET /v1/voices |
speechdash_get_voice | GET /v1/voices/{voice_id} |
speechdash_synthesize_speech | POST /v1/audio/speech |
Resource: speechdash://openapi (live OpenAPI JSON).
Streaming speech (POST /v1/audio/stream and POST /v1/audio/stream/with-timestamps) has no MCP tool. Use REST for those routes and for /health.
Setup: MCP server and the Agents hub.
Credits
Speech synthesis (REST and MCP) spends 0.5 display credits (1 wallet cent) per started 30 seconds of generated audio. Translate uses the same rate as MP3 export. Creating documents and file export do not spend credits. Creating documents counts toward your plan's daily document cap.
See What uses credits.