Granite 4.1 30B
Granite 4.1 30B is the instruction-tuned version of the matching Granite 4.1 base. It supports text generation, retrieval-based answers, extraction, coding and tool use.
Model design and training
IBM uses supervised fine-tuning and reinforcement-learning alignment. The training data combines public permissively licensed instructions, internal synthetic examples and selected human-authored data. The card names 12 supported languages and fill-in-the-middle code completion.Granite 4.1 30B model card
Practical use
Practical guidance. The 30B instruction release is suitable for comparing a general chat checkpoint with the newer reasoning model. Measure response quality and total generated tokens on the same requests. Keep tool execution outside the model and validate requested operations.
Specifications & access
| Detail | Value and conditions |
|---|---|
| Launch date | |
| Access | API, Downloadable, Provider product |
| Model identifiers | ibm-granite/granite-4.1-30b |
| Context window | 131,072 tokens Published model context; actual serving limits can be lower. Granite 4.1 30B model card |
| License | Apache 2.0Granite 4.1 30B model card |
| Parameters | 30BGranite 4.1 30B model card |
| Hosted access | watsonx: deploy on demand.IBM watsonx.ai foundation model library |
Related models
- Granite 4.2 30B
IBM · Granite 4.2
A 30B dense text model derived from Granite 4.1 Base with full, low-effort and direct-answer thinking modes.
Launch date Downloadable - Granite 4.1 30B Base
IBM · Granite 4.1
A pretrained 30B dense decoder for text generation, further training and fill-in-the-middle code completion.
Launch date Downloadable