Granite 4.2 30B
Granite 4.2 30B is an open-weight language model from IBM, released on August 25, 2026.
- Developer
- IBM
- Released
- August 25, 2026
- Total parameters
- 30B
- Architecture
- Dense
- Context window
- 128K
- Input
- Text
- License
- Apache 2.0
- Commercial use
- Yes
- Official weights, GB
- 58.5 GB
- Hugging Face repository
- huggingface.co
- Checked on
Granite 4.2 30B is a dense model with 30B parameters. IBM lists a context window of 128K tokens natively, with a long-context extension to 512K. The model takes text input. The weights are released under Apache 2.0, which allows commercial use. The official weights on Hugging Face take about 58.5 GB in BF16. Running the model needs at least that much memory across GPUs and system RAM, plus room for the context cache; quantized versions need less.
- Total parameters, billions
- 30B
- Context window, tokens
- 128K
- License type
- Apache 2.0
- License conditions
- Apache 2.0, with no user or revenue limits.
- Weights precision
- BF16
- Official quantized or alternative versions
- ibm-granite/granite-4.2-30b-fp8; ibm-granite/granite-4.2-30b-mxfp4; ibm-granite/granite-4.2-30b-nvfp4; ibm-granite/granite-4.2-30b-GGUF
- License text
- apache.org
- Sources
Page URL Hugging Face model card https://huggingface.co/ibm-granite/granite-4.2-30b/raw/main/README.md Hugging Face weights index https://huggingface.co/ibm-granite/granite-4.2-30b/raw/main/model.safetensors.index.json Hugging Face API https://huggingface.co/api/models?author=ibm-granite
Questions
What license does Granite 4.2 30B use?
How much memory does Granite 4.2 30B need?
What context window does Granite 4.2 30B have?
How does Granite 4.2 30B compare with Granite 4.2 8B?
Sources
Every figure comes from the developer’s model card, license text and official announcements, and from the model’s Hugging Face repository metadata. Each model page lists its sources and the date they were checked; the register is rechecked every week.