Models · Model provider
inclusionAI
inclusionAI publishes Ling and Ring language models for self-hosted instruction following and reasoning.
+ FollowBuilder facts
- Ling 3.1 Flash. Ling-3.1-flash uses hybrid attention and a mixture of experts with 560 billion total and 25 billion active parameters, with a one-million-token context. Source ↗
- Smaller checkpoints. Ling-mini-2.0 offers MIT-licensed base and post-trained checkpoints. Its 32,000-token context extends to 128,000 tokens with YaRN. Source ↗
- Hosted access. The Ling-3.1-flash model card links to downloadable weights for self-hosting and to separate hosted access through OpenRouter. Source ↗
- Ring reasoning. Ring-2.6-1T is a reasoning checkpoint with a 128,000-token context, extendable to 256,000 tokens using YaRN. Source ↗
Sources checked
Publisher update ·
inclusionAI updates Ling-3.1-flash-1019 inference code ↗The checkpoint’s Transformers code skips training-only multi-token prediction during ordinary generation and avoids inference-cache updates on the training path.
Change timeline
0 storiesNo verified releases in this timeline yet.