Archive — CN — technode
Ant Group launches Ling-3.1-flash with 560 billion parameters
Ant Group’s InclusionAI has launched the Ling-3.1-flash language model featuring 560 billion parameters, designed for various applications including agent tasks and office software. Currently in a two-week free trial limited to 256,000 tokens, the model aims to enable a one million token context window and will be released as open source afterward.
Key points
- Ant Group's InclusionAI launched Ling-3.1-flash.
- The model has 560 billion parameters with 25 billion activated per token.
- It is intended for agent tasks, search, office software, and specialized applications.
- Designed for context windows of up to one million tokens.
- Current trial limits usage to 256,000 tokens for two weeks.
- Future plans include enabling the larger window and open-source release.
Why it matters
- As of 2025, AI trends show a major shift toward larger language models supporting diverse applications.
- Increased interest in open-source AI reflects a broader trend toward collaboration and innovation in tech.
- Similar recent advancements include Google and OpenAI releasing competing models, emphasizing the arms race in AI development.
The launch of Ling-3.1-flash by Ant Group positions it strategically within the rapidly evolving landscape of AI, showcasing advancements in language processing and setting the stage for future developments in open-source technologies.