Open studio

One service, from idea to audio

Your podcast, your way

Voice prepared text, clean up raw text, or create a single-speaker topic episode with no search, Quick search, or Deep research. Dated news stays a separate private product.

The maximum charge is locked before the run. Audio starts private.

News Podcast

New episode

Private
01

Prepared text

Speech synthesis only

02

Raw text

Text preparation + voice

03

Topic and length

Optional research + script + preparation + voice

Bounded quote

Calculated for your input

before work
♡ Feed & favorites
⌁ API v1
Private until you publish
Maximum charge locked before the run
Telegram and browser

Three user-podcast modes

Start with what you already have

Every mode uses the same library, voices, credits, and publication flow. Only script preparation changes.

01

Prepared text

Send a narration-ready script directly to speech synthesis, with no LLM editing.

Speech synthesis only
02

Raw text

An agent preserves meaning and prepares numbers, abbreviations, and pronunciation for the narrator.

Text preparation + voice
03

Topic and length

Create a 20, 30, or 60 minute episode, choose a shorter preset, or set a custom 3–60 minute duration. Use model knowledge, Quick search, or Deep research.

Optional research + script + preparation + voice

From choice to private audio

1

Choose a mode and voice

Text or topic, language, length, and an available voice.

2

Confirm the quote

The server shows the expected charge and locks a maximum before work starts.

3

Listen, then choose whether to publish

The finished episode appears in your private library.

Current news is a separate path

Research depth matches the product

Topic Podcasts can use bounded Quick or Deep research. Dated News remains strict, private, and separate from the public Podcast Feed.

Bounded research

Agents use configured search tools within a quoted budget; a source count is not presented as a quality score.

Files, URLs, and collections

Save private TXT, Markdown, PDF, DOCX, and public URLs as immutable versions, organize them in collections, and reuse accepted research without another search run.

Hot-news alerts

Choose notification categories. The service alerts you to events that pass the operator-configured interest threshold for that category.

Scheduled digest

Choose topic, period, days, and time — the episode starts automatically.

Custom topics

Save your own areas of interest for researched news.

Multiple voice providers

The catalog shows voices that can be priced for the selected language.

Telegram and Web App

Create, listen, and manage episodes in a familiar interface.

Community without losing privacy

Publish only what you want to share

A new episode is private. After publication, authenticated users can discover it in the feed and play its audio.

For integrations

You manage the key. The API voices prepared text

Key lifecycle and API invocation are separate steps. Raw text and topic creation stay in the user studio.

  1. 01

    Create a key in Settings

    Name the key, select scopes, and optionally set a credit quota.

  2. 02

    Copy the secret once

    The service stores only a safe representation. Replace a lost secret with a new key.

  3. 03

    Revoke access

    The list shows state and usage; revoke keeps content-free operation history.

  4. 04

    Call the versioned TTS API

    Get a quote, confirm a prepared-text operation, poll status, and download private audio.

Request shape
Authorization: Bearer <YOUR_API_KEY>
POST /api/v1/tts/quotes
{ title, text, voiceId, language }
prepared_text only
quote → create → status → private audio

Pay for a confirmed operation

Buy credits with Telegram Stars

Before work starts, the server shows an expected charge and a fixed whole-credit maximum that includes the selected stages and voice.

Quote first

  • Price is bound to your text or topic
  • Change the input, recalculate the quote
  • Confirmation reserves the maximum; success captures the final bounded charge; failure or cancellation releases the remainder

Stars packages

  • Pay inside Telegram
  • No subscription — top up when needed
  • One balance for studio and API
View balance

Internal provider cost and margin are not shown to the user.

Frequently asked questions

A topic Podcast can use no search, Quick search, or Deep research and may be published to the Podcast Feed. News requires current dated material, remains private, and is regenerated for freshness.

An episode starts private. You can publish it later and unpublish it at any time.

The quote includes every selected stage. Confirmation reserves the locked maximum; success captures the final bounded charge and releases unused Credits; failure or cancellation releases the remaining reserve.

Only you. Likes affect the public count; favorites remain private.

No. The versioned developer API accepts prepared text only. Topic and raw text are available in the user studio.

Revoke it and create a new one. The secret is displayed only once at creation.

Which podcast will you create first?

Open the studio in Telegram or your browser, choose a mode, and see the expected charge plus a locked maximum.