// IBM Granite · Text LLM
granite-3.2-8b-instruct: VRAM & local deployment
8.17B parameters. About 7.3 GB at 4-bit and 19.8 GB at BF16 with one 8K-token request.
released 2025-02-17licence: Commercial OKmemory figures are estimates
01Specs
| Publisher | IBM · ibm-granite |
|---|---|
| Family / series | IBM Granite · granite-3.2 |
| Type | Text LLM (text-generation) |
| Parameters | 8.17B Source: counted from official safetensors weights |
| Context length | 128K tokens (131,072)max_position_embeddings in config.json |
| Architecture | 40 layers · GQA |
| Released weights | BF16 · 16.3 GBdownload |
| Licence | Apache-2.0 · Commercial OKPermissive licence (Apache-2.0 / MIT / CC-BY / OpenMDW): commercial use allowed. |
| Released | 2025-02-17 (Hugging Face repo created) |
02How much VRAM it needs (estimate)
Total = weights + 10% + 0.5GB overhead + KV cache for one request (FP16 KV). More concurrent users multiply the KV column.
| Precision | Weights | Total @ 8K | Total @ 32K | Total @ 128K | Smallest GPU | Mac |
|---|---|---|---|---|---|---|
| BF16 / FP16 (original) | 16.3 GB | 19.8 GB | 23.8 GB | 39.9 GB | RTX 4090 24GB | Mac 36GB |
| FP8 / INT8 / Q8_0 | 8.7 GB | 11.4 GB | 15.4 GB | 31.5 GB | 16GB GPU | Mac 24GB |
| 4-bit (Q4_K_M / AWQ / GPTQ-Int4) | 5.0 GB | 7.3 GB | 11.3 GB | 27.4 GB | 12GB GPU | Mac 16GB |
KV cache per request: 8K 1.3 GB · 32K 5.4 GB · 128K 21.5 GB
Bits per weight: BF16 16, FP8 8.5, 4-bit 4.85. “Fits” = total ≤ 90% of memory; Mac GPU can use ~2/3 of unified memory (≤36GB) or ~3/4 (larger). For MoE, all experts must be in memory, so size by total parameters. Method →
03How to run it
Only runtimes the publisher or the runtime’s official library names for this model.
- Transformers — mentioned in the official model card
- Ollama — in the official Ollama library · granite3.2:8b
ollama run granite3.2:8b04Official repositories & quantised builds
1 official repo(s) from ibm-granite.
- granite-3.2-8b-instruct Instruct
05Other IBM Granite models
All 67 →| Model | Params | 4-bit total | Released |
|---|---|---|---|
| granite-3.2-2b-instruct · same series | 2.53B | 2.9 GB | 2025-02-17 |
| granite-switch-4.1-8b | 8.78B | 7.8 GB | 2026-10-14 |
| granite-switch-4.1-30b | 30B | 22.7 GB | 2026-10-14 |
| granite-speech-5.0-2b | 2.31B | 2.0 GB | 2026-10-07 |
| ensemble-forecasting-with-ibm-granite-time-series | — | — | 2026-10-01 |
| granite-speech-5.0-470m-turboctc-nc | 473M | 0.8 GB | 2026-08-17 |
| granite-4.2-3b | 3.66B | 3.6 GB | 2026-08-07 |
| granite-4.2-8b | 8.79B | 7.7 GB | 2026-08-07 |
| granite-4.2-30b | 29.3B | 22.2 GB | 2026-08-07 |
06Sources
Data checked 2025-04-17. Parameters, context and licence are copied from the official repository; memory figures are our estimates.