llama-index-llms-ibm

What it is and what it does

This package bridges LlamaIndex's LLM abstraction layer with IBM's watsonx.ai foundation models. It wraps the ibm-watsonx-ai SDK to let you call IBM models (like granite variants) using LlamaIndex's standard completion and chat interfaces, supporting both synchronous and streaming modes.

You instantiate a WatsonxLLM class with your IBM Cloud credentials, project ID, and model choice, then call .complete(), .chat(), .stream_complete(), or .stream_chat() just as you would with any other LlamaIndex LLM provider. The package handles authentication, parameter marshalling, and response parsing against the IBM API.

Use it for:

• Build a RAG pipeline using LlamaIndex with IBM's foundation models as the inference backend.
• Stream responses from IBM watsonx.ai models in a chat application via LlamaIndex.
• Experiment with different IBM granite models without rewriting your LlamaIndex application code.
• Use IBM watsonx.ai in a multi-provider LLM setup where LlamaIndex abstracts the provider differences.

Worth the install?

AI-flagged interpretation of the facts on this page — verify before relying

Integrates IBM watsonx.ai foundation models with LlamaIndex's LLM API, enabling you to use IBM's models for text completion and chat inference through a unified interface.

Yes, if you are already using LlamaIndex and need to integrate IBM watsonx.ai models. The package has low install friction, permissive licensing, and active maintenance. No known vulnerabilities. Install it as a straightforward provider plugin for LlamaIndex.

Install

llama-index-llms-ibm on PyPI

pip install llama-index-llms-ibm

uv add llama-index-llms-ibm

poetry add llama-index-llms-ibm

Installing llama-index-llms-ibm

Before you install

Low install friction with three runtime dependencies. Actively maintained as of the latest release.

License in practice

MIT license permits commercial and private use with minimal restrictions.

Quickstart

pip install llama-index-llms-ibm

from llama_index.llms.ibm import WatsonxLLM

watsonx_llm = WatsonxLLM(
    model_id="ibm/granite-4-h-small",
    url="YOUR_URL",
    project_id="YOUR_PROJECT_ID"
)
response = watsonx_llm.complete("What is generative AI?")

Requires IBM Cloud API key set as WATSONX_APIKEY environment variable and a valid watsonx.ai project or space ID.

Verify before relying

• Whether all IBM watsonx.ai foundation models are supported or only a subset.
• Performance characteristics and latency expectations for different model sizes.
• Rate limiting or quota behavior when calling IBM's API through this integration.

Package facts

Evidence: llama_index_llms_ibm-0.7.0.post1-py3-none-any.whl

Tags

More Artificial Intelligence packages

Further reading