Granite 4.0 H Small is an open-source language model by ibm-granite. Features: LLM, VRAM: 64.4GB, Context: 128K, License: apache-2.0.
Arxiv:0000.00000 Conversational Deploy:azure Endpoints compatible Granite-4.0 Granitemoehybrid Language Region:us Safetensors Sharded Tensorflow
Model Card on HF 🤗: https://huggingface.co/ibm-granite/granite-4.0-h-small
Granite 4.0 H Small Parameters and Internals
| LLM Name | Granite 4.0 H Small |
| Repository 🤗 | https://huggingface.co/ibm-granite/granite-4.0-h-small |
| Required VRAM | 64.4 GB |
| Updated | 2026-03-28 |
| Maintainer | ibm-granite |
| Model Type | granitemoehybrid |
| Model Files | |
| Model Architecture | GraniteMoeHybridForCausalLM |
| License | apache-2.0 |
| Context Length | 131072 |
| Model Max Length | 131072 |
| Transformers Version | 4.56.0 |
| Tokenizer Class | GPT2Tokenizer |
| Padding Token | <|pad|> |
| Vocabulary Size | 100352 |
| Torch Data Type | bfloat16 |
Quantized Models of the Granite 4.0 H Small
Model |
Likes |
Downloads |
VRAM |
|---|---|---|---|
| Granite 4.0 H Small FP8 | 4 | 529 | 33 GB |
| ...e 4.0 H Small Unsloth Bnb 4bit | 3 | 55 | 58 GB |
Best Alternatives to Granite 4.0 H Small
Best Alternatives |
Context / RAM |
Downloads |
Likes |
|---|---|---|---|
| Granite 4.0 Tiny Preview | 128K / 13.3 GB | 26003 | 182 |
| Granite 4.0 H Tiny 5bit MLX | 128K / 4.8 GB | 21 | 2 |
Note: green Score (e.g. "73.2") means that the model is better than ibm-granite/granite-4.0-h-small .
Rank the Granite 4.0 H Small Capabilities
🆘 Have you tried this model? Rate its performance. This feedback would greatly assist ML community in identifying the most suitable model for their needs. Your contribution really does make a difference! 🌟
What open-source LLMs or SLMs are you in search of? 52052 in total.
Original data from HuggingFace, OpenCompass and various public git repos.
Check out Ag3ntum — our secure, self-hosted AI agent for server management.
Release v20260328a