fastvlm needs about 2.0 GB of memory at Q4_K_M (a 1.4 GB file) with an 8K context window, so it fits on a 4 GB GPU. This repository has 36 quants, from 1.0 GB (IQ4_NL) to 9.4 GB (Q8_0).
FastVLM 0.5B – lightweight vision‑language model
FastVLM 0.5B is a vision‑language model based on the FastVLM architecture. It uses a hybrid vision encoder that outputs fewer tokens, reducing encoding time. The model is available in multiple GGUF quantizations.
Good at
- Fast encoding of high‑resolution images
- Efficient use of VRAM
- Supports 8k context length
Limits
- No vision support beyond image encoding
- License is apple‑amlr
- No explicit language support listed
Best for
- Image captioning on low‑resource GPUs
- Rapid visual question answering
- Embedding generation for image‑text retrieval
Who it’s for. Developers needing a small vision‑language model for edge devices or local inference. Suitable for projects that require quick image encoding and moderate context handling.
Quants & memory
| Quant | File | Est. memory | vs VRAM | Result | |
|---|---|---|---|---|---|
| IQ4_NL | 1.0 GB | 1.6 GB | Fits | Get file → | |
| Q4_0 | 1.0 GB | 1.6 GB | Fits | Get file → | |
| Q4_1 | 1.0 GB | 1.6 GB | Fits | Get file → | |
| Q5_0 | 1.0 GB | 1.7 GB | Fits | Get file → | |
| Q5_1 | 1.1 GB | 1.7 GB | Fits | Get file → | |
| Q8_0 | 1.2 GB | 1.8 GB | Fits | Get file → | |
| Q2_K | 1.3 GB | 2.0 GB | Fits | Get file → | |
| Q3_K_M | 1.4 GB | 2.0 GB | Fits | Get file → | |
| Q4_K_M | 1.4 GB | 2.0 GB | Fits | Get file → | |
| Q5_K_M | 1.4 GB | 2.0 GB | Fits | Get file → | |
| Q6_K | 1.4 GB | 2.1 GB | Fits | Get file → | |
| F16 | 1.5 GB | 2.2 GB | Fits | Get file → | |
| BF16 | 1.5 GB | 2.2 GB | Fits | Get file → | |
| Q2_K | 1.6 GB | 2.3 GB | Fits | Get file → | |
| IQ3_XXS | 1.7 GB | 2.4 GB | Fits | Get file → | |
| IQ3_S | 1.8 GB | 2.5 GB | Fits | Get file → | |
| Q3_K_M | 1.8 GB | 2.5 GB | Fits | Get file → | |
| IQ4_XS | 1.9 GB | 2.6 GB | Fits | Get file → | |
| IQ4_NL | 1.9 GB | 2.7 GB | Fits | Get file → | |
| Q4_0 | 1.9 GB | 2.7 GB | Fits | Get file → | |
| Q4_K_M | 1.9 GB | 2.7 GB | Fits | Get file → | |
| Q4_1 | 2.0 GB | 2.8 GB | Fits | Get file → | |
| Q5_0 | 2.1 GB | 2.9 GB | Fits | Get file → | |
| Q5_K_M | 2.1 GB | 2.9 GB | Fits | Get file → | |
| Q5_1 | 2.2 GB | 2.9 GB | Fits | Get file → | |
| Q6_K | 2.3 GB | 3.0 GB | Fits | Get file → | |
| Q8_0 | 2.6 GB | 3.4 GB | Fits | Get file → | |
| F32 | 3.0 GB | 3.9 GB | Fits | Get file → | |
| Q2_K | 4.6 GB | 5.7 GB | Fits | Get file → | |
| IQ3_XXS | 5.0 GB | 6.1 GB | Fits | Get file → | |
| IQ3_S | 5.3 GB | 6.4 GB | Fits | Get file → | |
| IQ4_XS | 5.9 GB | 7.2 GB | Fits | Get file → | |
| IQ4_NL | 6.2 GB | 7.4 GB | Tight | Get file → | |
| Q4_0 | 6.2 GB | 7.4 GB | Tight | Get file → | |
| Q6_K | 7.8 GB | 9.3 GB | Too big | Get file → | |
| Q8_0 | 9.4 GB | 11.0 GB | Too big | Get file → |
For 8 GB VRAM, use Q4_0 or IQ4_NL (file size 1 GB, memory ~1.6 GB). For 12 GB VRAM, Q5_0 or Q5_1 (1.05–2.1 GB, memory 1.7–2.9 GB). For 24 GB VRAM, Q8_0 (1.18–9.41 GB, memory 1.8–11 GB).
What GPU do you need for fastvlm?
These figures add the file size, an fp16 KV cache at an 8K context window and about 0.5 GB for the runtime. They are estimates, so leave a little headroom, and expect to use less context or a smaller quant if you are right at the limit.
- 4 GB VRAM: the best quant that fits is F32 (3.0 GB file, about 3.9 GB in use).
- 6 GB VRAM: the best quant that fits is Q2_K (4.6 GB file, about 5.7 GB in use).
- 8 GB VRAM: the best quant that fits is IQ4_NL (6.2 GB file, about 7.4 GB in use).
- 12 GB VRAM: the best quant that fits is Q8_0 (9.4 GB file, about 11.0 GB in use).
Which fastvlm quant should you choose?
Q4_K_M is the usual starting point for fastvlm: it balances file size and quality well. Here it is a 1.4 GB file.
The Q8_0 file is 1.2 GB, about 0.9 times the size of Q4_K_M. It keeps the weights closer to the original, which only matters if you have the memory to spare.
If memory is tight, the smallest file is IQ4_NL at 1.0 GB, with the quality trade-offs that come with fewer bits. The usual sweet spot between file size and quality, and the quant most people start with.
New to quant names? Read GGUF quantization explained.
Specifications
Architecture from the GGUF header
Source
Frequently asked questions about fastvlm
How much VRAM does fastvlm need?
About 2.0 GB at Q4_K_M with an 8K context window. That is the 1.4 GB file, plus the KV cache and roughly 0.5 GB of runtime overhead. Smaller quants need less and longer context needs more.
Can I run fastvlm on an 8 GB GPU?
Yes, with IQ4_NL: the 6.2 GB file needs about 7.4 GB at 8K context. Larger quants in this repository do not fit in 8 GB.
What is the file size of fastvlm GGUF?
It depends on the quant: IQ4_NL is 1.0 GB, Q4_0 is 1.0 GB, Q4_1 is 1.0 GB, Q5_0 is 1.0 GB, Q5_1 is 1.1 GB, Q8_0 is 1.2 GB and 30 more.
Where can I download fastvlm GGUF?
The files are on Hugging Face in the gguf-org/fastvlm-gguf repository. The download buttons on this page open the file on Hugging Face. ggufmodels.com does not host any model files.
What license does fastvlm use?
Hugging Face lists the license as apple-amlr. Licenses can differ between the quantized files and the original model, so check the repository and the base model before commercial use.
Can I run FastVLM on 8 GB VRAM?
Yes, the Q4_0 or IQ4_NL quantizations require about 1.6 GB of VRAM at 8k context.
Does FastVLM support multiple languages?
The model does not list specific language support; it is primarily a vision‑language model.