Ling 2.6 Flash

LLM

inclusionAI

Ling 2.6 Flash is a 104B-parameter (7.4B active) instruct model designed for agent workflows requiring fast responses, strong task execution, and high token efficiency, with a focus on coding, document processing, and lightweight reasoning.

Context tokens

262,144

Output tokens

8,192

Docs

Model documentation

Released

Apr 24, 2026

Schema

This model is optimized to produce concise outputs by design, so responses may be notably shorter than expected. It is particularly tuned for agentic and multi-step tool-use workflows.

Schema documentation

Capabilities

Tool use

Structured output

Supported tools

No tools enabled.

Pricing

Type	Credits	Units
Input	1.33	Credits per 1k tokens
Output	3.99	Credits per 1k tokens
Cache hit	0.27	Credits per 1k tokens

Variants