Granite
by IBM · www.ibm.com ↗
IBM's family of open-weight language models aimed at enterprise and code tasks.
open-weights llm enterprise
- Category
- Open Models
- Business model
- Open (open weights/source)
- Availability
- Global
- Launched
- 2023-09
- Record updated
- 2026-07-30
- Model family
- Granite
- License
- Apache-2.0
- Founded
- 1911
- Headquarters
- US
- Canonical URL
- https://globalaiproductindex.com/products/granite/
Overview
Granite is IBM's family of open-weight language models aimed at enterprise tasks, including instruction-following, code, time-series, and guardrail variants. The models are comparatively small and efficient, released under the Apache 2.0 license, and integrated into IBM's watsonx platform. Weights are free to use, with paid hosted access through watsonx and other clouds.
Key features
- Open-weight models under Apache 2.0
- Enterprise focus including RAG, code, and safety variants
- Compact sizes optimized for efficiency
- Available through IBM watsonx and open repositories
Use cases
- Enterprise assistants with permissive licensing requirements
- On-premises deployments where data control matters
- Fine-tuning small models for business workflows
Pricing
Granite weights are free under Apache 2.0; hosted inference through IBM watsonx or partner clouds is billed by usage.
Frequently asked questions
What is IBM Granite?
Granite is IBM's family of open-weight, enterprise-focused language models covering chat, code, and safety tasks.
What license do Granite models use?
Granite models are released under the permissive Apache 2.0 license, allowing commercial use.
Where can Granite models run?
They can be self-hosted from open repositories or consumed as paid hosted models on IBM watsonx.
Similar products
- Baichuan — Open-weight bilingual LLM family from Beijing-based Baichuan Intelligence, now focused on medical AI with its Baichuan-M reasoning models.
- Command R — Cohere's open-weight LLM family optimized for retrieval-augmented generation and tool use; succeeded by Command A as Cohere's flagship.
- DBRX — Open-weight mixture-of-experts LLM released by Databricks in 2024; retired from Databricks serving in 2025, though the weights remain available.
- DeepSeek — Open-weight LLM family and chat assistant from Hangzhou-based DeepSeek, known for strong reasoning models. The official DeepSeek-V4-Flash-0731 API brings a major agent-capability upgrade via post-training, with native Responses API support and Codex adaptation.
- Falcon — Family of open-weight large language models developed by the Technology Innovation Institute (TII).
- Gemma — Google's family of lightweight open-weight models derived from the same research as Gemini.
Sources
This record was last reviewed on 2026-07-30.
Machine-readable record: /api/products/granite.json · Spot an error? Suggest a correction