Add IBM Granite Models to Langflow

[HugoCatarin0]

Enhancement Proposal: Add IBM Granite Models to Langflow

With DataStax now being acquired by IBM (announced Feb 2025, expected to close Q2 2025), integrating IBM Granite models into Langflow presents a unique opportunity. The Granite family (e.g., Granite 3.0, 400M–8B parameters) offers lightweight, enterprise-focused foundation models optimized for coding, RAG, reasoning, and function calling—perfectly aligning with Langflow’s mission to simplify and accelerate AI app development.

Advantages for Langflow/DataStax:

Strengthens synergy with IBM’s watsonx ecosystem, enhancing Langflow’s role as a middleware powerhouse for generative AI workflows.
Leverages IBM’s open-source commitment (Apache 2.0-licensed Granite models) to boost Langflow’s appeal to enterprise users.
Positions Langflow as a go-to platform for IBM’s cutting-edge AI offerings post-acquisition.
Benefits for the Langflow Community:

Access to high-performance, scalable models (trained on 4.5T tokens, 116 coding languages) for prototyping and deploying RAG and multi-agent apps.
Expands Langflow’s model-agnostic flexibility, empowering developers with more options for enterprise-grade solutions.
Fosters collaboration with IBM’s open-source initiatives (e.g., Granite and InstructLab), enriching the community’s innovation potential.
I’d love to hear your thoughts on adding Granite model support—could this be a game-changer for Langflow’s future?