Ling 3.1 Flash
Ling 3.1 Flash is a hybrid reasoning mixture-of-experts model from inclusionAI, with 25B active parameters out of 560B total.
Specifications
| Lab | InclusionAI |
|---|---|
| Context window | 262.1K tokens |
| Input price | $0.00/1M |
| Output price | $0.00/1M |
| Release | Oct 2026 |
Strengths
- 560B MoE with 25B active params per forward pass
- Hybrid reasoning with tool-call support for agents
- 262K context with 32K max completion tokens
- Zero-cost launch pricing on OpenRouter
Weaknesses
- No independent benchmark scores published yet
- No open-weight release listed at launch
- Launch pricing is likely temporary
Best for
- Agentic tool-use workflows on a budget
- Long-context reasoning and document work
- Prototyping before 3.1 pricing settles
In Depth: Ling 3.1 Flash
Summary
Ling 3.1 Flash is an AI model from InclusionAI.
Released in Oct 2026. It takes text input and produces text output, with a context window of 262.1K tokens. Input pricing is $0.00 per 1M tokens and output is $0.00 per 1M tokens on OpenRouter.