
An author finishes a 90,000-word novel. A professional narrator quotes $5,000 and a three-week turnaround. The royalty projections do not justify the cost. The book stays text-only, and potential listeners never hear it.
AI text-to-speech changes that math entirely. The same book can now be narrated using a cloned or generated voice in a fraction of the time and cost. For publishers with deep backlists, the technology turns thousands of unrecorded titles into audio products without proportional investment.
CAMB.AI's Stories feature inside DubStudio converts your written text into a fully narrated audiobook using AI text-to-speech. The process handles voice selection, script editing, speed adjustment, and multilingual export from a single interface. Here is how to do it step by step.
Start by uploading your ebook in TXT or DOCX format. The Stories feature also supports image uploads and scanned documents through OCR (Optical Character Recognition), which extracts text from images or scanned pages automatically.
Make sure your file does not exceed the platform's character limit for smooth processing. Once uploaded, the system detects and prepares the content for narration.
After uploading, select the language your ebook is written in. Accurate language selection ensures the AI reads pronunciation, rhythm, and intonation correctly. CAMB.AI supports text-to-speech narration in 150+ languages, covering 99% of the world's speaking population.
You have three options for the narrator's voice:
The narrator's voice stays consistent across the entire audiobook, eliminating the session-to-session drift that occurs with traditional multi-day studio recordings.
You can add an optional description to provide context about the story, its theme, or its intended audience. A clear description helps organize your projects when managing multiple audiobooks.
Once your text file is uploaded, the language is selected, and the narrator's voice is chosen, click "Begin Creating a Story." The system splits your text into individual lines, each containing a single sentence, and prepares them for audio generation.
Before generating audio, review the split script for sentence flow, proper line breaks, and correct punctuation. Punctuation affects how the AI reads your text, so keeping periods and commas on the correct line avoids audio artifacts.
Once satisfied, click "Re-generate all stale/muted dialogues audio" in the floating toolbar. The system generates audio for every line. After the generation completes, you can:
You can also select multiple lines at once to regenerate audio or change the speaker in bulk, which saves time on longer manuscripts.
After the narration is complete in the source language, you can produce multilingual editions of the same audiobook. Select "Add Other Languages" and choose between two options:
The author's voice is preserved in every language through voice cloning, creating a personal connection with international listeners that a different narrator in each language cannot match.
Click "Generate Audio Tracks" when editing is complete. A dialog box appears where you click "Export," and once processing finishes, click "Download." You can download:
For a mastered output with professional post-processing, the platform provides a separate export option within the same project page.
The audiobook market continues to grow as listeners consume content during commutes, workouts, and downtime. Converting your ebook into an audiobook opens your content to an audience that prefers audio over reading on a screen.
Key advantages include:
CAMB.AI's MARS-Pro model achieves 0.87 WavLM speaker similarity on the MAMBA benchmark, delivering voice consistency and emotional expression across long-form content. For audiobooks that require director-level emotion control, MARS-Instruct (1.2B parameters) adjusts delivery style per passage based on the emotional context of the text.
Whether you are an independent author, publisher, or content creator, AI text-to-speech narration makes audiobook production accessible at any scale.
Whether you're a media professional or voice AI product developer, this newsletter is your go-to guide to everything in speech and localization tech.


