Skip to content
Local AI Wizard
Your machine
Theme

granite-vision-3.3-2b

IBM · ibm-granite/granite-vision-3.3-2b

Weights license: Apache 2.0 · code and components may use other terms

Reported by the publisher19 resultsIBM's model card · read 2026-10-25 · not ranked · see them
Memory needed~3 GBQ4 weights + working room, before the machine's reserve
Parameters3 BHub safetensors total
Repository updated2026-04-02A commit is not a release date

Quality

No comparable score

We track no public, comparable benchmark that covers this model yet. No score is inferred.

Reported by the publisher

19 results · IBM's model card · read 2026-10-25
ChartQA
0.87Document benchmarks
DocVQA
0.91Document benchmarks
TextVQA
0.80Document benchmarks
AI2D
0.77Document benchmarks
InfoVQA
0.68Document benchmarks
OCRBench
0.79Document benchmarks
LiveXiv VQA v2
0.61Document benchmarks
LiveXiv TQA v2
0.52Document benchmarks
Every reported result19
MMMU
0.37Other benchmarks
VQAv2
0.79Other benchmarks
RealWorldQA
0.63Other benchmarks
VizWiz VQA
0.62Other benchmarks
OK VQA
0.55Other benchmarks
Politics
8.0Safety Evaluation
Racial
8.1Safety Evaluation
Jailbreak
7.5Safety Evaluation
Mislead
8.0Safety Evaluation
Unsafe Images (Unsafe)
8.4Safety Evaluation
Safe Images with Unsafe Instructions
9.3Safety Evaluation

Values as printed on the card, with the publisher's own settings (card section: Generation: › Safety Evaluation); column “Granite-vision-3.3-2b”. Not comparable across publishers and not used in any ranking. Model card

Runs on

Fast memory holds it 118

The planning estimate fits the memory the model can use quickly: graphics card VRAM, or unified memory within the macOS GPU share.

Show all 118

Within memory, outside the fast part 1

Within the machine's memory, outside its fast part: on a graphics card the layers that do not fit run on the CPU from PC RAM (the listed RAM, or 32 GB assumed); on a Mac, above the macOS default GPU share. Not a speed figure.

Hardware profiles grouped by the Q4 planning estimate on each machine (dedicated GPUs reserve 2 GB; unified and system memory 25%, 2 to 8 GB): the fast memory holds it, it is within memory but outside the fast part (layers offloaded to PC RAM, or above the macOS GPU share), or the machine is one memory tier short. Not a measured fit or speed.

Consider it for

IBM released granite-vision-3.3-2b on 2025-06-11: an image-to-text model, with 3B parameters on the Hugging Face count and the Apache 2.0 license. Languages the publisher declares: en.

Check first

License: commercial use allowed. At Q4 the weights load about 2.3 GB before each machine's reserve; image resolution, video frames and context add memory, and the vision encoder needs support in the runtime you pick.

Measured speed

No published measurement for this checkpoint

Tokens per second are shown only when a source publishes them with hardware, build and setup. They are never estimated.

How to run

Runtime compatibility and tokens per second have not been measured for this entry. Published metadata and a memory estimate do not prove a successful local run.

Documented ways to run it

Copied as printed in the linked source (read 2026-10-18); not run by this site.

Setup stated first by the source · Transformers
pip install transformers>=4.49
Transformers · model card · Usage with `transformers`
from transformers import AutoProcessor, AutoModelForVision2Seq
from huggingface_hub import hf_hub_download
import torch

device = "cuda" if torch.cuda.is_available() else "cpu"

model_path = "ibm-granite/granite-vision-3.3-2b"
processor = AutoProcessor.from_pretrained(model_path)
model = AutoModelForVision2Seq.from_pretrained(model_path).to(device)

# prepare image and text prompt, using the appropriate prompt template

img_path = hf_hub_download(repo_id=model_path, filename='example.png')

conversation = [
    {
        "role": "user",
        "content": [
            {"type": "image", "url": img_path},
            {"type": "text", "text": "What is the highest scoring model on ChartQA and what is its score?"},
        ],
    },
]
inputs = processor.apply_chat_template(
    conversation,
    add_generation_prompt=True,
    tokenize=True,
    return_dict=True,
    return_tensors="pt"
).to(device)


# autoregressively complete prompt
output = model.generate(**inputs, max_new_tokens=100)
print(processor.decode(output[0], skip_special_tokens=True))
Setup stated first by the source · vLLM
pip install torch torchvision torchaudio
pip install vllm==0.6.6
vLLM · model card · Usage with vLLM
from vllm import LLM, SamplingParams
from vllm.assets.image import ImageAsset
from huggingface_hub import hf_hub_download
from PIL import Image

model_path = "ibm-granite/granite-vision-3.3-2b"

model = LLM(
    model=model_path,
)

sampling_params = SamplingParams(
    temperature=0.2,
    max_tokens=64,
)

# Define the question we want to answer and format the prompt
image_token = "<image>"
system_prompt = "<|system|>\nA chat between a curious user and an artificial intelligence assistant. The assistant gives helpful, detailed, and polite answers to the user's questions.\n"

question = "What is the highest scoring model on ChartQA and what is its score?"
prompt = f"{system_prompt}<|user|>\n{image_token}\n{question}\n<|assistant|>\n"
img_path = hf_hub_download(repo_id=model_path, filename='example.png')
image = Image.open(img_path).convert("RGB")
print(image)

# Build the inputs to vLLM; the image is passed as `multi_modal_data`.
inputs = {
    "prompt": prompt,
    "multi_modal_data": {
        "image": image,
    }
}

outputs = model.generate(inputs, sampling_params=sampling_params)
print(f"Generated text: {outputs[0].outputs[0].text}")

Choose my computer

Technical details

What the source and the catalog store for this checkpoint. Missing fields are left out, never zero.

Task
image-to-text
Architecture
LlavaNextForConditionalGeneration
Weight license
Apache 2.0ibm-granite/granite-vision-3.3-2b model card (Hugging Face) · read 2026-10-18source
Commercial use
Allowed Apache License 2.0 text (SPDX) · read 2026-10-18license text
Languages
en Publisher's statement · read 2026-10-18source
Released
granite-vision-3.3-2b model card (Hugging Face, IBM) · read 2026-10-18source
Every stored field10 fields
Revision checked
bf66a8244401a8e26d2ff429aa82df8550a289e5
Architecture
LlavaNextForConditionalGeneration
Hub task tag
image-to-text
Formats
safetensors
Access
Open download
Publisher kind
Official
Repository created
2025-06-03creation is not the model's release
Repository updated
2026-04-02a commit is not a release date
Hub downloads (30 days) · likes
195,462 · 88this repository only · read 2026-10-25
Catalog state
Hub config or weights found; no inventory benchmark link

Sources

Identity and published specifications come from the checkpoint page. Benchmark scores belong to the named release and do not measure speed on your computer.

Curated facts checked: 2026-10-18. The license and release date cite the publisher's sources; parameters, files and task come from the Hugging Face record, and memory is a labelled estimate from its parameter count. A value no source states stays empty; a repository commit is not a model release date.