← The Polylog AI Intelligence Brief
Morning Edition · Tuesday, August 18, 2026Published at 2:24 AM EDT · New York
MiniMax Releases Open Weights for a Music Model That Generates Five-Minute Songs With Vocals
Music 3 pairs an 8-billion-parameter model initialized from Qwen3-8B with a 0.6-billion-parameter acoustic model, outputs 32 kHz stereo, and permits commercial use.
MiniMax published downloadable weights for Music 3, a model that takes lyrics and a structured description and returns a complete song with vocals, up to five minutes long, as 32 kHz 16-bit stereo audio. The company's release post describes…
Continue the AI Intelligence Brief
Track frontier labs, chips, export controls, model releases, regulation, and AI infrastructure.
- 5 AI intelligence signals a day
- Frontier labs, compute, and chips
- Model releases and AI infrastructure
- Source-grounded analysis with confidence labels
The Global Intelligence Brief stays free.
Townsquare
Start a discussion in Townsquare.
More from this edition
- Anthropic Tells Investors Its Revenue Run Rate Passed $65 Billion in July
- OpenAI Dissolves the Preparedness Team That Assessed Catastrophic Model Risks
- OpenAI Says It Is Training Models to Write Secure Code After Its Agent Breached Hugging Face
- Nvidia Says It Is Locking Up Land, Power and Shell Capacity Exclusively for Its Own AI Factories
- Cohere Labs Ships a 2.4-Billion-Parameter Vision Model That Reads Images at Native Resolution
- The Cheapest Listed Price for OpenAI's Top GPT-5.6 Model Has Halved Since Launch
- Paper Finds Agent Retries Inflate the True Cost of a Model Up to 4.25 Times Its Listed Price
Comments
0No comments yet.