MiniMax
MiniMax develops multimodal AI models for text and image tasks.
About MiniMax
MiniMax builds large-scale text and multimodal models optimized for generation and reasoning.
Notable offerings include models such as MiniMax-Text-01 available via third-party routers.
Company facts
- Founded
- 2021
- Headquarters
- Beijing, China
- Country
- People's Republic of China
Founders: Yan Junjie
Facts from Wikidata.
MiniMax products on Dhanasvi
Live listings of MiniMax's AI products across our directories.
Models
MiniMax M3
Processes long multimodal sequences across text, images, and video.
MiniMax M1
Processes million-token contexts for deep text analysis.
MiniMax-01
Processes over one million tokens of text and images in a single context.
MiniMax M2.7
MiniMax M2.7 handles massive text contexts up to 204800 tokens.
MiniMax M2
MiniMax M2 processes up to 204800 tokens of text in a single context.
MiniMax M2.5
MiniMax M2.5 processes up to 204800 tokens for extended text tasks.
MiniMax M2.1
MiniMax M2.1 handles massive contexts as a closed-source text LLM.
MiniMax M2-her
A closed-source LLM built for extended text processing and analysis.
Focus & what's next
- →Large language model development
- →Multimodal generation capabilities
- →Proprietary model APIs
- →Platform integrations for inference
Frequently asked questions
MiniMax is known for its proprietary multimodal and text AI models.
Related companies
Profile compiled from public sources and linked to live products by Dhanasvi's autonomous agents.