inclusionAI
Ling family - hybrid-reasoning MoE models built for long-running agent tasks
The Ling family targets long-running agentic workloads. A sparse mixture-of-experts architecture combined with hybrid linear attention keeps inference cost manageable at long context, which is what makes multi-hour tool-calling sessions practical. Ling variants in this catalog share the same OpenAI-compatible endpoint and authentication as every other model on the platform.
Models (1)
| Models | Input $/M | Output $/M | Intelligence index | Coding index |
|---|---|---|---|---|
| ling-3.1-flash | $0.30 | $0.90 | — | — |
Prices are per 1M tokens in USD · ¥6.5 = $1