Granite 4.0 Tiny Preview — Hardware Requirements & GPU Compatibility
ChatFunctions
Granite 4.0 Tiny Preview is a 6.7B-parameter open language model from IBM in the Granite family. It supports a context window of up to 131,072 tokens. At Q4_K_M it needs about 4.47 GB of VRAM — see which GPUs and Macs can run it below.
Runs great on a 24 GB GPU
Runs from 8 GB (Q4_K_M · 32K)Best on 24 GB (Q4_K_M · ~276 tok/s · 128K)Macs from 8 GB (Q4_K_M · ~48 tok/s)
Apache 2.0
Reference GPU · RTX 4090 24 GB
Estimates, ±20% · how we estimate
Run Granite 4.0 Tiny Preview
One recommended way. Every other method, OS and quant is in the step-by-step guide.Each way opens a step-by-step guide for your OS and quant.
Recommended
ollama run hf.co/bartowski/ibm-granite_granite-4.0-tiny-preview-GGUF:Q4_K_MQ4_K_M fits fully on the RTX 4090 · 18.4 GB free · 16K context
Download: ibm-granite_granite-4.0-tiny-preview-Q4_K_M.gguf4.12 GBbartowski
Ways to runOther ways
each opens step-by-stepRun in cloud
Runs on RTX 3060 12GB (6 GB headroom) · Q4_K_M
- Generation speed
- ~162 tok/s
- generation speed
- Cost per 1M output tokens
- $0.11
- per 1M output tokens
How Much VRAM Does Granite 4.0 Tiny Preview Need?
Select a quantization level; the listed hardware adheres to it.
| Quantization | VRAM | +32K | On RTX 4090 (reference) | |
|---|---|---|---|---|
| 3.3 GB | 5.8 GB | 2-bit · K-quant — 2-bit quantization with K-quant improvements | Fits: 128K context · ~308 tok/s | |
| 3.4 GB | 5.9 GB | 3-bit small — 3-bit small quantization | Fits: 128K context · ~305 tok/s | |
| 3.7 GB | 6.2 GB | 3-bit medium — 3-bit medium quantization | Fits: 128K context · ~296 tok/s | |
| 3.8 GB | 6.3 GB | 4-bit · legacy — 4-bit legacy quantization | Fits: 128K context · ~294 tok/s | |
| 4.5 GB | 7 GB | 4-bit medium — 4-bit medium quantization — most popular sweet spot | Fits: 128K context · ~276 tok/s | |
| 5.2 GB | 7.7 GB | 5-bit medium — 5-bit medium quantization — good quality/size tradeoff | Fits: 128K context · ~259 tok/s | |
| 6.0 GB | 8.5 GB | 6-bit, very good — 6-bit quantization, very good quality | Fits: 128K context · ~244 tok/s | |
| 7.1 GB | 9.7 GB | 8-bit, near-lossless — 8-bit quantization, near-lossless | Fits: 128K context · ~224 tok/s |
Selected Q4_K_M
4-bit medium quantization5.6GBat 16K context
- Weights
- 4.0 GB
- KV cache
- 1.3 GB
- Overhead
- 0.3 GB
Fits RTX 4090 (reference) · 18.4 GB free · ~276 tok/s
≥50 tok/s needs ≥80 GB/s · RTX 5090 ~334, RTX 3090 Ti ~276
Which Hardware Can Run Granite 4.0 Tiny Preview?
Q4_K_M · 4.5 GBGranite 4.0 Tiny Preview (Q4_K_M) requires 4.5 GB of VRAM to load the model weights. For comfortable inference with headroom for KV cache and system overhead, 6+ GB is recommended. Using the full 131K context window can add up to 10.6 GB, bringing total usage to 15.0 GB. 52 GPUs can run it, including NVIDIA GeForce RTX 5090, NVIDIA GeForce RTX 3090 Ti. 163 devices with unified memory can run Granite 4.0 Tiny Preview, including NVIDIA DGX Spark, NVIDIA Jetson AGX Thor Developer Kit, MacBook Air 13" M3 (8 GB).
≤ 11 GB
≤ 11 GB tier →fits · 2.4 GB free · up to 32K ctx
16 GPUs · 1 Mac class · 2 device families
RTX 3080~245Simulate output on RTX 3080RTX 2080 Ti~221Simulate output on RTX 2080 TiRTX 3070 Ti~220Simulate output on RTX 3070 TiM3*8 GB~48Simulate output on iMac 24" M3 (8 GB)Orin Nano~62Simulate output on NVIDIA Jetson Orin Nano 8GB (Super)iPhone~36Simulate output on iPhone 1720 GB
fits · 14.4 GB free · up to 128K ctx
1 GPU
RX 7900 XT~101Simulate output on RX 7900 XT32–40 GB
fits · 26.4 GB free · up to 128K ctx
1 GPU · 3 Mac classes · 4 device families
RTX 5090~334Simulate output on RTX 5090M5 Max36 GB~92Simulate output on Mac Studio M5 Max (36 GB)M532 GB~60Simulate output on MacBook Air 13" M5 (32 GB)Orin~109Simulate output on NVIDIA Jetson AGX Orin 32GBSnapdragon X Elite (X1E-80-100)~69Simulate output on Snapdragon X Elite Copilot+ PC
By VRAM tier · typical GPUs and Macs at Q4_K_M
Alternatives to Granite 4.0 Tiny Preview
Granite family · 28 models →Similar models
6| Model | Score | VRAM | tok/s | Why it is similar |
|---|---|---|---|---|
| Llava 1.5 7B HFllava-hf | — | 5 GB | ~141 | same size |
| Chatglm3 6BZ.ai | — | 5.2 GB | ~159 | same size |
| Baichuan2 7B Basebaichuan-inc | — | 5 GB | ~142 | same size |
| Internlm 7BInternLM | — | 15.4 GB | ~43 | same size |
| Xgen 7B 8k BaseSalesforce | — | 5.7 GB | ~142 | same size |
| RedPajama INCITE 7B Basetogethercomputer | — | 4.6 GB | ~142 | same size |
Specifications
- Publisher
- IBM
- Family
- Granite
- Parameters
- 6.7B
- Active parameters
- 1.4B per token
- Architecture
- GraniteMoeHybridForCausalLM
- Context Length
- 131,072 tokens
- Vocabulary Size
- 49,160
- Release Date
- 2025-04-30
- License
- Apache 2.0
Frequently Asked Questions
How much VRAM does Granite 4.0 Tiny Preview need?
Granite 4.0 Tiny Preview requires 4.5 GB of VRAM at Q4_K_M, or 13.8 GB at BF16. Full 131K context adds up to 10.6 GB (15.0 GB total).
VRAM = Weights + KV Cache + Overhead
Weights = 6.7B × 4.8 bits ÷ 8 = 4 GB
KV Cache + Overhead ≈ 0.5 GB (at 2K context + ~0.3 GB framework)
Fit ratings and hardware model lists check this model with room for a 16K-token context, which needs a little more memory.
KV Cache + Overhead ≈ 11 GB (at full 131K context)
VRAM usage by quantization
What's the best quantization for Granite 4.0 Tiny Preview?
For Granite 4.0 Tiny Preview, Q4_K_M (4.5 GB) offers the best balance of quality and VRAM usage. Q4_K_L (4.5 GB) provides better quality if you have the VRAM. The smallest option is IQ2_M at 2.7 GB.
VRAM requirement by quantization
★ Recommended — best balance of quality and VRAM usage.
Can I run Granite 4.0 Tiny Preview on a Mac?
Yes — MacBook Air 13" M3 (8 GB) and 145 other Macs can run Granite 4.0 Tiny Preview. Apple Silicon uses unified memory, so the model shares RAM with the system. At Q4_K_M you need at least 4.5 GB of usable unified memory (RAM minus macOS overhead).
Can I run Granite 4.0 Tiny Preview locally?
Yes — Granite 4.0 Tiny Preview can run locally on consumer hardware. Popular tools include Ollama, LM Studio, and llama.cpp.
How fast is Granite 4.0 Tiny Preview?
At Q4_K_M, Granite 4.0 Tiny Preview can reach ~122 tok/s on AMD Instinct MI350X. On NVIDIA GeForce RTX 4090: ~276 tok/s. Speed depends mainly on GPU memory bandwidth. Real-world results typically within ±20%.
tok/s = 1000 ÷ (active GB ÷ (bandwidth GB/s × efficiency) × 1000 + layers × routing ms)
Mixture-of-Experts: just the active experts get read for each token, along with a fixed routing cost per layer.
Example: NVIDIA B200 → 0.9 GB active ÷ (8000 × 0.65) = 0.18 ms, plus 40 layers × 0.055 ms = 2.20 ms, so 1000 ÷ 2.38 ms = ~420 tok/s
Estimated speed at Q4_K_M (4.5 GB)
Real-world results typically within ±20%. Speed depends on batch size, quantization kernel, and software stack.
What's the download size of Granite 4.0 Tiny Preview?
At Q4_K_M, the download is about 4.00 GB. The full-precision BF16 version is 13.34 GB. The smallest option (IQ2_M) is 2.25 GB.
Which GPUs can run Granite 4.0 Tiny Preview?
52 consumer GPUs can run Granite 4.0 Tiny Preview at Q4_K_M. Top options include AMD Radeon RX 6700 XT, AMD Radeon RX 6800, AMD Radeon RX 6800 XT. 52 GPUs have plenty of headroom for comfortable inference.
Which devices can run Granite 4.0 Tiny Preview?
163 devices with unified memory can run Granite 4.0 Tiny Preview at Q4_K_M (4.5 GB), including AMD Ryzen AI 9 HX 370 (Strix Point) Laptop, ASUS Ascent GX10, Beelink GTR9 Pro (Ryzen AI Max+ 395, 128 GB), Framework Desktop (Ryzen AI Max+ 395, 128 GB). Apple Silicon Macs use unified memory shared between CPU and GPU, making them well-suited for local LLM inference.