A text-to-podcast API converts written material into a spoken episode. Dayfaro takes source text, generates an editable script, and records it with synthetic voices. The result is an MP3 and a transcript that you can use in your own application. You can run the same workflow in the Dayfaro workspace or through its MCP server.
Start with material you already have: an article you can use, research notes, a newsletter draft, or a product update. Dayfaro’s podcast endpoint works from the text you send. It does not open a URL, research a topic, or decide which news is current. Those steps belong in your application before the generation request.
What do you send to the podcast endpoint?
Create a Dayfaro account, add prepaid credits, and generate an API key in your profile. Send a POST request to /v1/podcast with your text, language, speaker count, voice choices, and target duration. Keep the API key on your server rather than exposing it in browser code.
For a first episode, use one speaker, English, the Kore voice, and a target of 60 seconds. Set generate_audio to true for the full script-and-audio workflow. The response identifies a background job; it is not the finished audio file. The API quickstart includes a complete request and the following status and download steps.
How do you know when the audio is ready?
Store the returned job ID and poll its status. When the job succeeds, retrieve the podcast and its download information. If the job fails, inspect the reported error before submitting another generation request. Repeating a paid generation blindly can create work you did not intend.
Use a stable idempotency key for a logical request. This helps your application handle a lost response without creating another episode. Keep the original identifier when retrying the same submission; use a new one when the user intentionally requests different content. The TypeScript SDK is another way to submit and follow this workflow.
Can you approve the script before recording?
Yes. Set generate_audio to false, inspect the script, and make your edits before requesting audio. This is useful for pronunciation, names, figures, and wording that needs approval. A new voice recording can use the existing script, so changing the voice does not require rewriting your source.
Dayfaro supports one to four speakers and six languages: English, Italian, Spanish, French, German, and Portuguese. Target durations run from 20 seconds to 10 minutes. Treat the target as guidance; script length, pauses, and delivery affect the final recording.
What should you handle before publishing?
Listen to the result and compare important claims with the source. Keep attribution alongside the episode where appropriate, and use only material you have permission to process and distribute. Generated speech should not be presented as a recording of a real person.
Download links expire, so retrieve the file while its link is valid or request a fresh link through your authenticated account. Your prepaid balance works across the API, MCP, and workspace. Purchases begin at $10, with no subscription and no expiry date for unused credits.