- Home
- Skills
- Agents & Orchestration
- Sarvam Studio
Works with the AI Agents you already use
Sarvam Studio
Transform topics or PDFs into multi-speaker AI podcasts with Sarvam-105B and Bulbul-v3 voice synthesis.
$5
Sarvam Studio
Example session with this skill installed
Make a medium length podcast in Hindi about the future of UPI using report.pdf as a source. Use Aayan and Kavya as hosts with tech_upbeat music.
- Read your context and instructions
- Compiled the sarvam studio
I have generated the Hindi podcast based on the provided UPI report. The script was grounded in the PDF content, featuring a 6-turn dialogue between Aayan and Kavya with an upbeat tech background track.
Podcast Generation Results
| Attribute | Details |
|---|---|
| Topic | भारत में फिनटेक और UPI का भविष्य |
| Status | Completed Successfully |
| Language | hi-IN (Hindi) |
| Host 1 | Aayan (Voice: aayan) |
| Host 2 | Kavya (Voice: kavya) |
| BGM Style | tech_upbeat |
| Duration | 3m 54s |
| Audio File | output/podcast_upi_future_1715604201.wav |
| Transcript | output/podcast_upi_future_1715604201_transcript.md |
| Metadata | output/podcast_upi_future_1715604201_transcript.json |
Next steps
- Review the generated Markdown transcript in the
outputfolder for content accuracy. - Listen to the
.wavfile to verify voice modulation and background music ducking. - If the episode is too short, re-run the command with the
--turns 10flag.
Connects securely to your tools. The creator never sees your data.
About this skill
Generating broadcast-quality audio content requires complex orchestration between LLMs, TTS engines, and audio post-production. This skill handles the entire pipeline to transform raw topics or documents into polished, multi-speaker podcasts tailored for the Indian market.
What it does
- Script generation uses Sarvam-105B to write natural two-host dialogue in English, Hindi, or Hinglish.
- Realistic voice synthesis assigns distinct personalities using Bulbul-v3 voices for diverse, human-like narration.
- Audio post-production automatically stitches dialogue segments with background music and ducking.
- Document grounding parses PDFs or text files to ensure the conversation is factually accurate to your source material.
- Multi-format exports delivers a production-ready
.wavfile along with Markdown and JSON transcripts.
How it works
- Provide a topic or a path to a source document (PDF/TXT) to ground the discussion.
- Select a language (en-IN or hi-IN) and configure the episode length via the turns parameter.
- Choose host voices and background music styles like
lofi_ambientortech_upbeat. - Execute the Python script to generate the script, synthesize audio, and assemble the final episode.
Frameworks & tools
Built with Python 3.10+, utilizing the Sarvam AI API (Sarvam-105B and Bulbul-v3). Requires pypdf for document processing and environment-based API key management.
Why this beats prompting it yourself
Manual prompting for podcasts often results in single-voice monologues or repetitive dialogue patterns. This skill manages the specific timing, multi-voice switching, and background music layering that basic LLM prompts cannot handle alone.
Use cases
- Convert internal technical reports into engaging audio briefings for team members.
- Create localized educational content in Hindi from English textbooks or notes.
- Generate quick daily news summaries with a professional radio-host aesthetic.
- Prototype podcast scripts and audio segments for marketing campaigns.
Known limitations
Requires a Sarvam API key and a stable internet connection. Text-based PDFs are supported, but scanned or image-based PDFs will fail to parse.
Find more skills like this on the Claude Skills Marketplace, the Codex Skills Marketplace and the Gemini Skills Marketplace, or browse Claude Skills for Agents & Orchestration.
Requires Sarvam API key. Supports text-based PDFs only (no OCR for scanned images). Script JSON parsing may occasionally require a retry. Python 3.10+ required.
How to use
- 1
Connect your agent to the Agensi MCP
Copy the prompt and paste it into Claude, ChatGPT, Cursor or any agent that supports MCP. It sets up the connection for you. Manual setup
- 2
Sign in to Agensi
Your agent shows a sign-in link. Sign in and click Allow.
- 3
Ask for what you need
Your agent loads this skill and uses it on your task. It always asks before buying a skill.
Reviews
No reviews yet
Be one of the first to try it. Every listed skill goes through the trust checks below.
Security scanned
Checked by our 8-point scan before listing
Fresh listing
Recently published to Agensi
Money-back guarantee
Doesn't do what its page promises? You get your money back.
Trust & safety
Security scan
Security scan passed 6 days ago
- Accounts
Sarvam
Permissions
Allowed Hosts
Needs network access to call the Sarvam API for script generation. Needs file read access to parse the source document, and write access to save the final .wav and transcript files."