Ling 3.1 Flash

Ling 3.1 Flash is a language model from InclusionAI for coding, analysis and agents that use software tools. It can work through a problem before returning an answer. Vercel added the model to AI Gateway on 30 September 2026, with free access through 13 October 2026.

Hosted access provides a 262K-token context

Vercel lists 560 billion total parameters, with 25 billion active per token, and a 262K-token context window. The context is the space available for instructions, documents, conversation history and the response. It is the useful limit to check when planning a long task.

The model uses a mixture-of-experts design, which selects part of its trained network for each piece of text. A large total parameter count and a smaller active count describe different aspects of its size.

Provider dates and prices can differ

OpenRouter’s listing, checked on 5 October, gives 2 October 2026 as its release date and offers a free NovitaAI route with a 262K context. That is a later provider listing than Vercel’s announcement. The release date on this page uses the earlier verified availability.

The free offer is introductory. Check the selected provider’s price, context limit and terms when the task runs. These listings establish access through hosted services; each provider controls its own availability and limits.