MiniMax M3 Open-Sourced, Response Speed Doubles
MiniMax, a prominent AI company based in China, today announced the open-source release of its native multimodal flagship model, MiniMax M3. Prior to this, the team had already released the model's weights to the industry and published a highly technical paper on sparse attention mechanisms, which drew significant attention from the open-source community.

Outstanding Performance Leads the Open-Source Race
As MiniMax's flagship model, M3 features a total of 428 billion parameters, with 23 billion active. Notably, it is the first open-source large language model to undergo multimodal mixed training from the very start of its training process.
Thanks to its advanced architecture, M3 rapidly claimed the top spot among open-source models on a leading global intelligence index within two weeks of launch. Following an initial surge in traffic that caused delays, the team optimized the model's output speed from 30 TPS to 80 TPS, with further enhancements planned.
Top-Tier Coding and Code Delivery Capabilities
In challenging coding and agent capability benchmarks, MiniMax M3 has achieved industry-leading performance. It can autonomously break down complex tasks and skillfully leverage external tools for multi-step reasoning.
Compared to traditional models, M3's code generation ability has taken a qualitative leap. The team stated that its code is designed to be deliverable-ready, not a barely functional prototype that needs heavy manual edits—a shift that will significantly improve developer productivity.
Related article
ByteDance’s Seed launches global campus drive, offering virtual shares to win top large model talent
In the competitive landscape of large language models, securing top-tier talent remains the most critical strategic asset.On April 1st, ByteDance announced the launch of its Seed global campus recruitment initiative, part of its large model talent de
Suno to Watermark Songs Amid Legal Battles
Suno, the platform enabling users to generate AI-created music, has unveiled new features to label platform-produced tracks, restrict downloads, and update community standards to curb unauthorized replicas. These updates arrive as Suno confronts mult
Musk Admits Grok Build Leaked User Code, Promises to Erase All Historical Data
Elon Musk directly addressed the privacy controversy surrounding Grok Build, beginning with a simple "True" to confirm the incident's validity. He pledged that all user data previously uploaded to SpaceXAI would be permanently erased, stating, "not a
Related Special Topic Recommendations
Comments (0)
0/500
MiniMax, a prominent AI company based in China, today announced the open-source release of its native multimodal flagship model, MiniMax M3. Prior to this, the team had already released the model's weights to the industry and published a highly technical paper on sparse attention mechanisms, which drew significant attention from the open-source community.

Outstanding Performance Leads the Open-Source Race
As MiniMax's flagship model, M3 features a total of 428 billion parameters, with 23 billion active. Notably, it is the first open-source large language model to undergo multimodal mixed training from the very start of its training process.
Thanks to its advanced architecture, M3 rapidly claimed the top spot among open-source models on a leading global intelligence index within two weeks of launch. Following an initial surge in traffic that caused delays, the team optimized the model's output speed from 30 TPS to 80 TPS, with further enhancements planned.
Top-Tier Coding and Code Delivery Capabilities
In challenging coding and agent capability benchmarks, MiniMax M3 has achieved industry-leading performance. It can autonomously break down complex tasks and skillfully leverage external tools for multi-step reasoning.
Compared to traditional models, M3's code generation ability has taken a qualitative leap. The team stated that its code is designed to be deliverable-ready, not a barely functional prototype that needs heavy manual edits—a shift that will significantly improve developer productivity.
ByteDance’s Seed launches global campus drive, offering virtual shares to win top large model talent
In the competitive landscape of large language models, securing top-tier talent remains the most critical strategic asset.On April 1st, ByteDance announced the launch of its Seed global campus recruitment initiative, part of its large model talent de
Suno to Watermark Songs Amid Legal Battles
Suno, the platform enabling users to generate AI-created music, has unveiled new features to label platform-produced tracks, restrict downloads, and update community standards to curb unauthorized replicas. These updates arrive as Suno confronts mult
Musk Admits Grok Build Leaked User Code, Promises to Erase All Historical Data
Elon Musk directly addressed the privacy controversy surrounding Grok Build, beginning with a simple "True" to confirm the incident's validity. He pledged that all user data previously uploaded to SpaceXAI would be permanently erased, stating, "not a





Home






