aifollow.news
Search
蚂蚁百灵宣布开源 Ling-3.1-flash，模型最高支持 1M 上下文，并加强了工具调用对 JSON Schema 的支持。在对应的 Hopper GPU 测试配置中，量化使权重占用从约 534GB 降至 292GB，服务部署由 8 卡降至 4 卡；实际所需配置仍取决于上下文、并发量等条件。
Coverage of the same occurrence shares a node; subsequent developments have their own nodes.
Ant Ling announced the release of Ling-3.1-flash as an open-source model. It supports up to 1M context and has improved JSON Schema support for tool calls. In the specified Hopper GPU test configuration, quantization reduced weight storage from about 534GB to 292GB and the serving setup from eight GPUs to four; actual requirements still vary with context length, concurrency and service configuration.
Ling-3.1-flash has approximately 560B total parameters and activates about 25B per token. Ant Ling says its context window supports up to 1M, but the planned two-week free trial will limit service length to 256K. The team plans to offer 1M context and release the model as open source after the service becomes paid.
By original publication time, with each report's bookmarks and feedback preserved.
Ant Ling announced the release of Ling-3.1-flash as an open-source model. It supports up to 1M context and has improved JSON Schema support for tool calls. In the specified Hopper GPU test configuration, quantization reduced weight storage from about 534GB to 292GB and the serving setup from eight GPUs to four; actual requirements still vary with context length, concurrency and service configuration.
Ling-3.1-flash has approximately 560B total parameters and activates about 25B per token. Ant Ling says its context window supports up to 1M, but the planned two-week free trial will limit service length to 256K. The team plans to offer 1M context and release the model as open source after the service becomes paid.
同一事件，精选展示《蚂蚁百灵开源 Ling-3.1-flash，最高支持 1M 上下文》