

MiniMax
β Text source-checked Nov 7, 2026MiniMax develops a family of proprietary foundation models across text and multimodal capabilities.
Source-checked against the vendor's site Β· rebuilt from the vendor's own pagesProfile updated Link checked How we rate
In short
MiniMax offers proprietary foundation models for text, speech, and multimodal tasks. It provides AI-native products like MiniMax Code and Design for developers and enterprises.
MiniMax is committed to advancing the frontiers of artificial intelligence and realizing Artificial General Intelligence (AGI). MiniMax develops a family of proprietary foundation models across text and multimodal capabilities. MiniMax offers AI-native products including MiniMax Code and MiniMax Design, alongside an open platform for enterprises and developers worldwide.
The MiniMax M3.1 Flash model is positioned for agent performance and cost efficiency. The MiniMax H3 model is an open model designed to break boundaries between tasks and modalities. MiniMax Music 3.0 is an open-weights model for music generation.
MiniMax Code is a product designed to minimize effort and maximize intelligence. MiniMax Design is a multimodal AI agent that orchestrates image, video, voiceover, and editing models to create content from idea to finished cut.
The MiniMax AI Platform is available on Web, Windows, Mac, iOS, and Android. The MiniMax API Platform supports online payment and offline bank transfer for account top-ups.
According to MiniMax: MiniMax is committed to advancing the frontiers of artificial intelligence and realizing Artificial General Intelligence (AGI). Its pages list this strength: MiniMax provides foundation models with multimodal capabilities. They also state this limitation: Voice cloning services require individual or enterprise real-name verification.
β The AI Tool Bible editorial team
Pros
- β MiniMax provides foundation models with multimodal capabilities.
- β The MiniMax H3 video model is an open model.
- β MiniMax Music 3.0 is an open-weights model.
- β MiniMax offers a suite of AI-native products and an open platform for enterprises and developers.
- β The platform is available across Web, Windows, Mac, iOS, and Android.
Cons
- β οΈ Voice cloning services require individual or enterprise real-name verification.
Facts
- Platforms
- web, windows, macos, ios, android
- API available
- Yes
- Free trial
- Yes (not source-checked)
- Company
- MiniMax (not source-checked)
- Launched
- Jan 1, 2022 (not source-checked)
- Model used
- MiniMax M3, Hailuo 2.3, Speech 2.8, Music 2.6 (not source-checked)
- Hacker News
- 11 stories Β· top 518 points Β· last Oct 19, 2026 Β· checked Nov 16, 2026
- Earliest Wayback capture
- 2021-12-17 (earliest archived capture of the site, not a launch date)
- Headquarters
- People's Republic of China (Wikidata)
- OpenRouter
- largest-context model minimax/minimax-m3 Β· 1,048,576 tokens Β· prompt price 0.0000003 per token (as listed) Β· checked Nov 16, 2026
Where this tool sits
Use cases
Frequently asked
- How much does MiniMax speech synthesis cost?
- The speech-2.8-hd (Sync) model costs $100 per million characters. This pricing applies to the paid tier of the MiniMax platform for speech generation services.
- What is the price for image generation?
- The image-01 model costs $0.0035 per image. This is part of the paid pricing structure for MiniMax's multimodal capabilities.
- Which platforms support the MiniMax AI Platform?
- The MiniMax AI Platform is available on Web, Windows, Mac, iOS, and Android. The API Platform supports online payment and offline bank transfer for account top-ups.
- What specific products does MiniMax offer?
- MiniMax offers AI-native products including MiniMax Code and MiniMax Design. MiniMax Code minimizes effort, while Design orchestrates image, video, voiceover, and editing models.
- What models are available in the MiniMax family?
- The family includes MiniMax M3, Hailuo 2.3, Speech 2.8, and Music 2.6. Specific models like M3.1 Flash focus on agent performance, while H3 is an open model.