Audio & Speech
Sep 14, 2026

The Gist
ElevenLabs launched Music v2.5 and, with Universal Music Group, a platform for remixing catalog songs, while OpenAI opened its GPT-Live-1 speech API at $0.05 per minute. Pocket FM also hit a $500 million run rate with AI generating 99% of its new content. AI is now cheap and good enough to produce and reshape music and speech at scale.
Today's Stories
- 1
ElevenLabs debuts Music v2.5, preferred in 47,885-pair blind test
ElevenLabs released Music v2.5 for ElevenMusic, also available via API, with free tier (five lossless downloads per day) and Pro tier (400 per month). Listeners preferred v2.5 most of the time in a 47,885-pair blind test, and the company says it trained its models on licensed stems and music.
The Universal Music Group licensing deal applies only to future products, not Music v2.5, so the test is whether that deal eventually covers the core product.
- 2
UMG, ElevenLabs launch catalog remix platform
Universal Music Group announced Thursday a partnership with voice AI firm ElevenLabs, letting users draw on its music catalog — artists must opt in — to create remixes and new takes on tracks. The deal, following AI music giant Suno's model with Warner Music and BMG, shifts AI music firms from lawsuit targets to paying customers; UMG, Sony and Warner sued Suno in 2024.
The bet is that letting fans play with the tech may demystify AI music, though artists must opt in for their work to be used. Watch whether more artists join in.
- 3
OpenAI opens GPT-Live-1 speech API at $0.05 per minute
OpenAI is making GPT-Live-1, a full-duplex speech model that listens and talks at once, available to developers as an API at $0.05 per minute. Yelp already uses it for phone reservations. On OpenAI's benchmarks, full-duplex interactivity rose to 80.1 percent from 45.4 percent for GPT-Realtime-2.1, while turn-taking latency fell to 0.8 seconds from 1.4 seconds.
Whether the $0.05 per minute price holds back adoption for high-volume calling, and whether external testing matches OpenAI's own benchmark gains. Watch the tool-calling accuracy improvement, 87 percent from 60 percent.
- 4
Pocket FM hits $500M run rate as AI powers 99% of new content
Pocket FM doubled its annualized revenue run rate to $500 million in a year, CEO Rohan Nayak said. AI now powers 93% of its catalog and 99% of new content. AI has made Pocket FM's content production about 80x cheaper, Nayak said, and 100 hours that once took a year can now be made in a day. Its 12-month revenue retention rate rose to 76% from 44% two years ago.
The growth hinges on whether Pocket Entertainment's AI-driven model extends beyond audio, as its three-month-old microdrama app Pocket Saga nears a $15 million run rate. Watch the reported $100 million to $120 million raise at about a $2 billion valuation.
- 5
Universal Music Group and ElevenLabs launch AI music platform
Universal Music Group announced Thursday it is launching an AI-powered platform with ElevenLabs, under a multiyear licensing agreement, letting users create remixes, mashups, and new takes on tracks from its licensed catalog. The deal adds to UMG's existing AI licensing arrangements with Spotify, Nvidia, and Klay, and an AI music platform it is developing with Udio. The new platform stays separate from ElevenLabs' Music API and ElevenMusic generator.
Whether artists opt in will shape how much of UMG's catalog actually fuels the platform. The press release lists no timeline or pricing for the announced "additional products and fan experiences in the months and years to come."
- 6
GPT-Live-1 Launches on API at $0.05 per Minute
OpenAI released GPT-Live-1 on its API. It's a single model that listens and speaks simultaneously, with the front-end voice layer priced at $0.05 per minute. Instruction-following has improved, and it supports custom voices and phone-based interactions. It replaces the STT-LLM-TTS chained setup, which introduces latency and is prone to interruptions.
Whether developers adopt the single-model approach for phone handling and customer support will be the litmus test. Using custom voices requires contacting sales.
What to Watch
For Universal Music Group’s deal with the AI music platform, watch whether it eventually covers the core product, not just future offerings, and whether enough artists opt in to make the catalog meaningful. Also keep an eye on whether the $0.05-per-minute price and sales-gated custom voices hold back adoption, and whether Pocket Entertainment can extend its AI-driven model beyond audio.
Sources
- Elevenlabs makes Music v2.5 available via app and API with free and pro tier options
- UMG and voice AI giant ElevenLabs launch AI-powered music platform
- OpenAI's GPT-Live-1 API lets developers build apps that talk and listen at the same time
- India’s Pocket FM doubles revenue run rate to $500M as AI powers 93% of audio content
- Universal Music is launching an AI music platform with ElevenLabs
- Build more natural voice experiences with GPT‑Live‑1 in the API
- Suno releases its first AI music model made with record industry help
- Read the Apple document explaining how new listening features still protect your privacy
- Suno launches v6 music models built with Warner, BMG, and Believe
- Amazon Prime Video’s new AI tech matches lips to dubbed audio
- Suno replaces its AI models with a new one trained on licensed music as copyright suits pile up
- Caterpillar’s Digital Chief Ogi Redzic on How A.I. Is Redefining Construction Work
- 蔦谷好位置とNovel Coreが語る『チート』、クラシックのサンプリングから生成AIまで
- 三菱電機、「タスク汎用音源分離技術」開発…単一AIモデルで複数の音源分離を実現 1枚目の写真・画像
- Meta's new real-time audio model is the foundation for AI assistants that never stop listening
- Google brings AI music generation directly into the Gemini app with its new Lyria 3.5 model
Share this with a friend
Send today's roundup to anyone who wants to keep up.
Get daily AI news free with AIToday
Today's AI, for Today's work. Email / LINE / Slack.
Sign up free