MODELS
inclusionAI: Ling 3.0 Flash Fin
by InclusionAI
Overview
A finance-enhanced Ling-3.0-flash model from InclusionAI, released with BF16 weights under the MIT license.
Details
Ling-3.0-flash-Fin is Ant Ling’s finance-enhanced model built from Ling-3.0-flash. The official Hugging Face model card publishes BF16 weights under the MIT license. Its Ling-3.0-flash base model is documented as a hybrid reasoning mixture-of-experts architecture with 124B total parameters and 5.1B active parameters. OpenRouter also offers a free variant with a listed 262,144-token context window and function-calling support.
When to Use
Use for finance-oriented language and reasoning workloads. Use when you need downloadable BF16 model weights under the MIT license. Evaluate it against the general Ling-3.0-flash model when finance specialization may affect task performance.
Getting Started
- Open the official Ling-3.0-flash-Fin model card on Hugging Face.
- Review the Ling-3.0-flash base-model card for architecture and inference guidance.
- Use the documented SGLang or vLLM quickstarts for the base model.
- Evaluate outputs on representative finance inputs before production use.
Key Features
- •Finance-enhanced variant of Ling-3.0-flash
- •BF16 weights
- •MIT license
- •Built from a hybrid reasoning mixture-of-experts base model
- •Base model: 124B total parameters and 5.1B active parameters
- •Base-model inference quickstarts for SGLang and vLLM
- •Free OpenRouter variant with listed function-calling support
Capabilities
- •Finance-focused language modeling
- •Reasoning
- •Mixture-of-experts inference
- •Function calling through the listed OpenRouter variant
Last updated Sep 7, 2026