Turn a script into speech from your conversation. Connect Crikk to ChatGPT, Claude, or another MCP client to find voices and create downloadable MP3 audio.
Use your existing Crikk API key and API allowance.
From a prompt to an MP3
Example request
Use Crikk to find an English voice, turn this sentence into speech, and give me the MP3 download: “Welcome to our next chapter.”
Start here
MCP lets an AI assistant use Crikk’s speech tools on your behalf. Before connecting, open API Tokens to get your API key and numeric user ID. You’ll need a Crikk account with API access and a client that supports custom remote MCP servers.
MCP server URL
https://crikk.com/mcp Transport: Streamable HTTP. Paste this URL into your client’s server URL field.
Connect with OAuth
Connect with OAuth
Enter your API key only in Crikk’s authorization page or your client’s credential settings. With OAuth, the client receives separate access tokens. Never paste your key into a chat prompt.
If your client supports custom request headers, use the same server URL and send both headers below. In Claude, choose No sign in and add these under Request headers. Replace the placeholders with your own credentials.
Authorization: Bearer YOUR_CRIKK_API_KEY
X-Crikk-User-Id: YOUR_USER_IDOn shared Claude connectors, members use the same configured API credentials and allowance. Choose OAuth for individual Crikk accounts. ChatGPT uses the OAuth flow above.
After connecting, paste this prompt into a conversation with Crikk enabled.
Use Crikk to find an English voice, turn this sentence into speech, and give me the MP3 download: “Welcome to our next chapter.”
You can also ask for a different language, a faster reading speed, or a speaking style supported by the selected voice.
Your assistant can choose a voice, start generation, and retrieve the result for you.
list_voices Browse available voices, languages, and speaking styles. Filter by a language or locale such as English or en-US.
Inputs: language (optional)
generate_speech Send your text with a voice name from the voice list. Adjust speed, pitch, or a supported speaking style. Returns a job ID.
Inputs: text, voice_name; optional speaking_style, speed, pitch
get_speech_result Check the same job until it completes, then download the audio. Repeated checks of a completed job return the same link without another charge.
Inputs: job_id
API and MCP access includes only Crikk’s free voice catalog. Premium voices are not available through these integrations. Speech generation uses your account’s API character allowance.
Yes. Use your existing API key and numeric Crikk user ID. Both are available on the API Tokens page. New keys are 24 characters long; existing shorter keys remain valid.
No. Speech generated through MCP uses the same API character allowance as your direct API requests. Characters are charged once when the completed result is retrieved. Check your account’s API allowance before generating.
Ask your assistant to check the existing job again. Do not submit the same text as a new generation while it is pending. Retrieve pending jobs within ten minutes; expired or failed jobs require a new generation.
Check the server URL, user ID, and API key. For OAuth, use automatic client registration rather than published client identity (CIMD). Your ChatGPT or Claude plan and workspace permissions must allow custom connections; a workspace administrator may need to enable them.
Audio links expire one hour after the result is finalized. Download and save the MP3 before then. Anyone with the link can download it during that hour, so share it only with people who should have access.
Remove the connection in your AI app. To invalidate access tied to a Crikk key, delete that key from the API Tokens page. This also stops any other integrations using the same key. OAuth connections may require you to reconnect after 30 days.