Sourced through our studio and distributor network · Cleared in writing

Licensed
podcast audio
for voice AI.

We partner with podcast studios, hosts, and distributors to license existing catalogs and commission new recordings for AI training — every speaker consented in writing.

Request a sample → Browse the catalog →
Built for the teams training speech recognition, text-to-speech, voice cloning, and conversational LLMs. Talk to the founder — every conversation starts with a sample.
How we work
Sourcing model
Network
StudiosPartner network
SpeakersWritten …consent
SourcingTo your spec
LicensePer project
What we source
Podcast audio
Custom
SourceStudio podcasts
SpeakersWritten consent
FormatWAV + transcripts
LicensePer project
Sample · EN-US · 48kHz
Two-speaker conversational
Cleared
00:00 / 00:60 · listen to a real catalog clip
or download the WAV
48 kHz24-bit WAV
−23 LUFSbroadcast
Diarized+ JSON
Built for the teams shipping the next generation of voice AI
[ Customer logo ][ Customer logo ][ Customer logo ][ Customer logo ][ Customer logo ][ Customer logo ] [ Customer logo ][ Customer logo ][ Customer logo ][ Customer logo ][ Customer logo ][ Customer logo ]
Cleared
in writing
Consent model
Every speaker signs a release that names AI training
Studio
grade
Audio quality
Recorded on broadcast chains in treated rooms
Native
verified
Multilingual
Sourced through a global network of podcast studios
Auditable
end‑to‑end
Provenance
Every file traceable to a named, contactable speaker
§ 01 — Catalog

Audio we can source
for your training use case.

We work with podcast studios and distributors to source and commission audio that matches the model you're training. Tell us the use case — we'll send a representative sample.

Solution · 01

ASR training data

Multi-speaker conversational audio with word-level alignment and diarization, sourced through our partner studios.

  • Aligned, diarized transcripts
  • Real-room acoustic diversity
  • Long-form conversational context
  • Cleared for commercial training
Explore ASR data →
Solution · 02

TTS training data

Phonetically balanced studio reads from named speakers, recorded on broadcast-grade chains for high-fidelity neural TTS and vocoder work.

  • Phoneme + prosody tags
  • Pronunciation lexicon
  • Broadcast-grade loudness
  • Cleared for commercial training
Explore TTS data →
Solution · 03

Voice cloning data

Single-speaker sets we commission with cloning rights named in the speaker release itself.

  • Cloning rights in writing
  • Identity-verified speakers
  • Multi-take, multi-emotion
  • Consent in writing
Explore cloning data →
Solution · 04

Conversational AI

Long-form, multi-turn dialogue with disfluencies, overlap, and turn-taking preserved — what audio-in/audio-out LLMs actually need to learn from.

  • Multi-turn dialogue
  • Turn-taking preserved
  • Backchannels intact
  • Register diversity
Explore conversational →
Solution · 05

Multilingual speech

A growing catalog of languages and regional locales, sourced through native-speaker podcast networks across multiple continents.

  • Native-speaker verified
  • Per-language transcripts
  • Regional accent coverage
  • New locales on request
Explore languages →
Solution · 06

Custom commission

We brief, cast, and record to your spec — domain, accents, scenarios, hours. You own the result, optionally on an exclusive license.

  • Scoped to brief
  • Exclusive licensing available
  • Direct studio access
  • Same provenance pipeline
Scope a commission →
§ 02 — Packages

Ways to
work with us.

We license existing podcast catalogs, source new audio through our studio network, and commission custom recordings to spec. Every speaker we record signs a written release.

Package 01

Conversational Core

Two- and three-speaker English podcast dialogue, studio-recorded, broad domain coverage across our partner network.
  • EN-US, EN-GB, EN-AU coverage
  • Studio-recorded, cleared
  • Diarized JSON + WAV
  • Licence per project
Licensed per project · request a quote
Package 02

Multilingual Expansion

A growing set of languages and regional locales, sourced through native-speaker podcast networks worldwide.
  • Multilingual, multi-locale
  • Native speaker verified
  • Per-language transcripts
  • Licence per project
Licensed per project · request a quote
Package 03

Custom Commission

We brief, cast, and record to your spec — domain, accents, scenarios, hours. You own the result.
  • Scoped to brief
  • 4–8 week turnaround
  • Exclusive licensing available
  • Direct studio access
Scoped to brief · request a quote
provenance.json — aipodcast
// what every recording we commission carries
{
  "file": "en-us/conv-core/0421/04221.wav",
  "consent": "signed_release",
  "duration_s": 187.42,
  "sample_rate": 48000,
  "speakers": ["S-01122", "S-01147"],
  "recording_id": "REC-04221",
  "speaker_named": true,
  "license": "per_project",
  "recorded_at": "2026-03-14T11:02:00Z"
}
§ 03 — Provenance

Every recording tied
to a named speaker.

Every recording we commission ties back to a written speaker release. We name a human contact at our company on every deal so your legal team has someone to call.

  • 01Written speaker release on file for every recording we commission
  • 02Recording details: speaker name, date, and consent version
  • 03Named human contact — speakers retain ownership and we honor it
§ 04 — Compliance

Built so your legal team
says yes the first time.

Every speaker we record signs a written release before we hit record. No scraped audio. No platform terms-of-service ambiguity.

 
aipodcast
Scraped web audio
User-generated platforms
Written speaker consent
Yes — every speaker
No
Implied / ToS
Commercial training rights
Negotiated per project
Unclear
Often prohibited
Signed release on file
Yes — written
No
No
Studio audio quality
48 kHz / 24-bit
Variable
Variable
§ 05 — Who you're working with
Jaeden Schafer

Jaeden Schafer, founder

I run Podcast Studio AZ and host the AI Chat podcast (top 10 tech in the US). aipodcast is how I'm turning a network of working podcasters into a clean, consented audio source for the teams training voice AI.

Every conversation starts with me. Email partnerships@aipodcast.io and I'll send a representative sample.

Founder-led
Every conversation starts with me
Cleared in writing
Written release from every speaker we record
Studio network
Sourcing through working podcasters across genres
§ 06 — Sample pack

Request a
sample.

Tell us what you're training and we'll get back to you with a representative sample and a short note on how we can source what you need.

  • 01You fill out the form (under 60 seconds)
  • 02We get back to you quickly with a sample
  • 03If it fits, we set up a 20-min call with the founder
§ 07 — Get started

Clean training data,
on a contract you can show your lawyer.

Every project starts with a sample and a quick scoping call with the founder. Licence terms are negotiated per project, with a named contact at our company on every deal.

Request a sample → hello@aipodcast.io