Skip to content
Local AI Wizard
Your machine
Theme

InternVL2-2B

OpenGVLab · OpenGVLab/InternVL2-2B

Weights license: MIT · code and components may use other terms

Reported by the publisher32 resultsOpenGVLab's model card · read 2026-10-25 · not ranked · see them
Memory needed~2 GBQ4 weights + working room, before the machine's reserve
Parameters2.2 BHub safetensors total
Repository updated2025-03-25A 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

32 results · OpenGVLab's model card · read 2026-10-25
DocVQA test
86.9InternVL2-2B · Image Benchmarks
ChartQA test
76.2InternVL2-2B · Image Benchmarks
InfoVQA test
58.9InternVL2-2B · Image Benchmarks
TextVQA val
73.4InternVL2-2B · Image Benchmarks
OCRBench
784InternVL2-2B · Image Benchmarks
MME sum
1,876.8InternVL2-2B · Image Benchmarks
RealWorldQA
57.3InternVL2-2B · Image Benchmarks
AI2D test
74.1InternVL2-2B · Image Benchmarks
Every reported result32
MMMU val
36.3InternVL2-2B · Image Benchmarks
MMBench-EN test
73.2InternVL2-2B · Image Benchmarks
MMBench-CN test
70.9InternVL2-2B · Image Benchmarks
CCBench dev
74.7InternVL2-2B · Image Benchmarks
MMVet GPT-4-0613
44.6InternVL2-2B · Image Benchmarks
MMVet GPT-4-Turbo
39.5InternVL2-2B · Image Benchmarks
SEED-Image
71.6InternVL2-2B · Image Benchmarks
HallBench avg
37.9InternVL2-2B · Image Benchmarks
MathVista testmini
46.3InternVL2-2B · Image Benchmarks
OpenCompass avg
54.0InternVL2-2B · Image Benchmarks
MVBench
60.2InternVL2-2B · Video Benchmarks
MMBench-Video 8f
0.97InternVL2-2B · Video Benchmarks
MMBench-Video 16f
1.03InternVL2-2B · Video Benchmarks
Video-MME w/o subs
45.0InternVL2-2B · Video Benchmarks
Video-MME w subs
47.3InternVL2-2B · Video Benchmarks
avg.
77.7InternVL2‑2B · Grounding Benchmarks
RefCOCO (val)
82.3InternVL2‑2B · Grounding Benchmarks
RefCOCO (testA)
88.2InternVL2‑2B · Grounding Benchmarks
RefCOCO (testB)
75.9InternVL2‑2B · Grounding Benchmarks
RefCOCO+ (val)
73.5InternVL2‑2B · Grounding Benchmarks
RefCOCO+ (testA)
82.8InternVL2‑2B · Grounding Benchmarks
RefCOCO+ (testB)
63.3InternVL2‑2B · Grounding Benchmarks
RefCOCO‑g (val)
77.6InternVL2‑2B · Grounding Benchmarks
RefCOCO‑g (test)
78.3InternVL2‑2B · Grounding Benchmarks

Values as printed on the card, with the publisher's own settings (card section: Performance › Image Benchmarks; Performance › Video Benchmarks). Not comparable across publishers and not used in any ranking. Model card

Runs on

Fast memory holds it 119

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

Show all 119

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

OpenGVLab released InternVL2-2B on 2024-07-04: a vision-language model that answers about images, with 2.21B parameters on the Hugging Face count and the MIT license. Languages the publisher declares: Multilingual.

Check first

License: commercial use allowed. At Q4 the weights load about 1.7 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.

Library in source: transformers

Documented ways to run it

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

Transformers · model card · Inference with Transformers
import numpy as np
import torch
import torchvision.transforms as T
from decord import VideoReader, cpu
from PIL import Image
from torchvision.transforms.functional import InterpolationMode
from transformers import AutoModel, AutoTokenizer

IMAGENET_MEAN = (0.485, 0.456, 0.406)
IMAGENET_STD = (0.229, 0.224, 0.225)

def build_transform(input_size):
    MEAN, STD = IMAGENET_MEAN, IMAGENET_STD
    transform = T.Compose([
        T.Lambda(lambda img: img.convert('RGB') if img.mode != 'RGB' else img),
        T.Resize((input_size, input_size), interpolation=InterpolationMode.BICUBIC),
        T.ToTensor(),
        T.Normalize(mean=MEAN, std=STD)
    ])
    return transform

def find_closest_aspect_ratio(aspect_ratio, target_ratios, width, height, image_size):
    best_ratio_diff = float('inf')
    best_ratio = (1, 1)
    area = width * height
    for ratio in target_ratios:
        target_aspect_ratio = ratio[0] / ratio[1]
        ratio_diff = abs(aspect_ratio - target_aspect_ratio)
        if ratio_diff < best_ratio_diff:
            best_ratio_diff = ratio_diff
            best_ratio = ratio
        elif ratio_diff == best_ratio_diff:
            if area > 0.5 * image_size * image_size * ratio[0] * ratio[1]:
                best_ratio = ratio
    return best_ratio

def dynamic_preprocess(image, min_num=1, max_num=12, image_size=448, use_thumbnail=False):
    orig_width, orig_height = image.size
    aspect_ratio = orig_width / orig_height

    # calculate the existing image aspect ratio
    target_ratios = set(
        (i, j) for n in range(min_num, max_num + 1) for i in range(1, n + 1) for j in range(1, n + 1) if
        i * j <= max_num and i * j >= min_num)
    target_ratios = sorted(target_ratios, key=lambda x: x[0] * x[1])

    # find the closest aspect ratio to the target
    target_aspect_ratio = find_closest_aspect_ratio(
        aspect_ratio, target_ratios, orig_width, orig_height, image_size)

    # calculate the target width and height
    target_width = image_size * target_aspect_ratio[0]
    target_height = image_size * target_aspect_ratio[1]
    blocks = target_aspect_ratio[0] * target_aspect_ratio[1]

    # resize the image
    resized_img = image.resize((target_width, target_height))
    processed_images = []
    for i in range(blocks):
        box = (
            (i % (target_width // image_size)) * image_size,
            (i // (target_width // image_size)) * image_size,
            ((i % (target_width // image_size)) + 1) * image_size,
            ((i // (target_width // image_size)) + 1) * image_size
        )
        # split the image
        split_img = resized_img.crop(box)
        processed_images.append(split_img)
    assert len(processed_images) == blocks
    if use_thumbnail and len(processed_images) != 1:
        thumbnail_img = image.resize((image_size, image_size))
        processed_images.append(thumbnail_img)
    return processed_images

def load_image(image_file, input_size=448, max_num=12):
    image = Image.open(image_file).convert('RGB')
    transform = build_transform(input_size=input_size)
    images = dynamic_preprocess(image, image_size=input_size, use_thumbnail=True, max_num=max_num)
    pixel_values = [transform(image) for image in images]
    pixel_values = torch.stack(pixel_values)
    return pixel_values

# If you want to load a model using multiple GPUs, please refer to the `Multiple GPUs` section.
path = 'OpenGVLab/InternVL2-2B'
model = AutoModel.from_pretrained(
    path,
    torch_dtype=torch.bfloat16,
    low_cpu_mem_usage=True,
    use_flash_attn=True,
    trust_remote_code=True).eval().cuda()
tokenizer = AutoTokenizer.from_pretrained(path, trust_remote_code=True, use_fast=False)

# set the max number of tiles in `max_num`
pixel_values = load_image('./examples/image1.jpg', max_num=12).to(torch.bfloat16).cuda()
generation_config = dict(max_new_tokens=1024, do_sample=True)

# pure-text conversation (纯文本对话)
question = 'Hello, who are you?'
response, history = model.chat(tokenizer, None, question, generation_config, history=None, return_history=True)
print(f'User: {question}\nAssistant: {response}')

question = 'Can you tell me a story?'
response, history = model.chat(tokenizer, None, question, generation_config, history=history, return_history=True)
print(f'User: {question}\nAssistant: {response}')

# single-image single-round conversation (单图单轮对话)
question = '<image>\nPlease describe the image shortly.'
response = model.chat(tokenizer, pixel_values, question, generation_config)
print(f'User: {question}\nAssistant: {response}')

# single-image multi-round conversation (单图多轮对话)
question = '<image>\nPlease describe the image in detail.'
response, history = model.chat(tokenizer, pixel_values, question, generation_config, history=None, return_history=True)
print(f'User: {question}\nAssistant: {response}')

question = 'Please write a poem according to the image.'
response, history = model.chat(tokenizer, pixel_values, question, generation_config, history=history, return_history=True)
print(f'User: {question}\nAssistant: {response}')

# multi-image multi-round conversation, combined images (多图多轮对话,拼接图像)
pixel_values1 = load_image('./examples/image1.jpg', max_num=12).to(torch.bfloat16).cuda()
pixel_values2 = load_image('./examples/image2.jpg', max_num=12).to(torch.bfloat16).cuda()
pixel_values = torch.cat((pixel_values1, pixel_values2), dim=0)

question = '<image>\nDescribe the two images in detail.'
response, history = model.chat(tokenizer, pixel_values, question, generation_config,
                               history=None, return_history=True)
print(f'User: {question}\nAssistant: {response}')

question = 'What are the similarities and differences between these two images.'
response, history = model.chat(tokenizer, pixel_values, question, generation_config,
                               history=history, return_history=True)
print(f'User: {question}\nAssistant: {response}')

# multi-image multi-round conversation, separate images (多图多轮对话,独立图像)
pixel_values1 = load_image('./examples/image1.jpg', max_num=12).to(torch.bfloat16).cuda()
pixel_values2 = load_image('./examples/image2.jpg', max_num=12).to(torch.bfloat16).cuda()
pixel_values = torch.cat((pixel_values1, pixel_values2), dim=0)
num_patches_list = [pixel_values1.size(0), pixel_values2.size(0)]

question = 'Image-1: <image>\nImage-2: <image>\nDescribe the two images in detail.'
response, history = model.chat(tokenizer, pixel_values, question, generation_config,
                               num_patches_list=num_patches_list,
                               history=None, return_history=True)
print(f'User: {question}\nAssistant: {response}')

question = 'What are the similarities and differences between these two images.'
response, history = model.chat(tokenizer, pixel_values, question, generation_config,
                               num_patches_list=num_patches_list,
                               history=history, return_history=True)
print(f'User: {question}\nAssistant: {response}')

# batch inference, single image per sample (单图批处理)
pixel_values1 = load_image('./examples/image1.jpg', max_num=12).to(torch.bfloat16).cuda()
pixel_values2 = load_image('./examples/image2.jpg', max_num=12).to(torch.bfloat16).cuda()
num_patches_list = [pixel_values1.size(0), pixel_values2.size(0)]
pixel_values = torch.cat((pixel_values1, pixel_values2), dim=0)

questions = ['<image>\nDescribe the image in detail.'] * len(num_patches_list)
responses = model.batch_chat(tokenizer, pixel_values,
                             num_patches_list=num_patches_list,
                             questions=questions,
                             generation_config=generation_config)
for question, response in zip(questions, responses):
    print(f'User: {question}\nAssistant: {response}')

# video multi-round conversation (视频多轮对话)
def get_index(bound, fps, max_frame, first_idx=0, num_segments=32):
    if bound:
        start, end = bound[0], bound[1]
    else:
        start, end = -100000, 100000
    start_idx = max(first_idx, round(start * fps))
    end_idx = min(round(end * fps), max_frame)
    seg_size = float(end_idx - start_idx) / num_segments
    frame_indices = np.array([
        int(start_idx + (seg_size / 2) + np.round(seg_size * idx))
        for idx in range(num_segments)
    ])
    return frame_indices

def load_video(video_path, bound=None, input_size=448, max_num=1, num_segments=32):
    vr = VideoReader(video_path, ctx=cpu(0), num_threads=1)
    max_frame = len(vr) - 1
    fps = float(vr.get_avg_fps())

    pixel_values_list, num_patches_list = [], []
    transform = build_transform(input_size=input_size)
    frame_indices = get_index(bound, fps, max_frame, first_idx=0, num_segments=num_segments)
    for frame_index in frame_indices:
        img = Image.fromarray(vr[frame_index].asnumpy()).convert('RGB')
        img = dynamic_preprocess(img, image_size=input_size, use_thumbnail=True, max_num=max_num)
        pixel_values = [transform(tile) for tile in img]
        pixel_values = torch.stack(pixel_values)
        num_patches_list.append(pixel_values.shape[0])
        pixel_values_list.append(pixel_values)
    pixel_values = torch.cat(pixel_values_list)
    return pixel_values, num_patches_list

video_path = './examples/red-panda.mp4'
pixel_values, num_patches_list = load_video(video_path, num_segments=8, max_num=1)
pixel_values = pixel_values.to(torch.bfloat16).cuda()
video_prefix = ''.join([f'Frame{i+1}: <image>\n' for i in range(len(num_patches_list))])
question = video_prefix + 'What is the red panda doing?'
# Frame1: <image>\nFrame2: <image>\n...\nFrame8: <image>\n{question}
response, history = model.chat(tokenizer, pixel_values, question, generation_config,
                               num_patches_list=num_patches_list, history=None, return_history=True)
print(f'User: {question}\nAssistant: {response}')

question = 'Describe this video in detail.'
response, history = model.chat(tokenizer, pixel_values, question, generation_config,
                               num_patches_list=num_patches_list, history=history, return_history=True)
print(f'User: {question}\nAssistant: {response}')
LMDeploy · model card · A 'Hello, world' Example
from lmdeploy import pipeline, TurbomindEngineConfig
from lmdeploy.vl import load_image

model = 'OpenGVLab/InternVL2-2B'
image = load_image('https://raw.githubusercontent.com/open-mmlab/mmdeploy/main/tests/data/tiger.jpeg')
pipe = pipeline(model, backend_config=TurbomindEngineConfig(session_len=8192))
response = pipe(('describe this image', image))
print(response.text)

Choose my computer

Technical details

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

Checkpoint / source
OpenGVLab/InternVL2-2B
Task
image-text-to-text
Architecture
InternVLChatModel
Library in source
transformers
Weight license
MITOpenGVLab/InternVL2-2B model card (Hugging Face) · read 2026-10-18source
Commercial use
Allowed MIT license text (SPDX) · read 2026-10-18license text
Languages
MultilingualModel card metadata · read 2026-10-25source
Released
InternVL README news (GitHub) · read 2026-10-18source
Every stored field13 fields
Revision checked
e4f6747bd20f139e637642c6a058c6bd00b36919
Architecture
InternVLChatModel
Library
transformers
Hub task tag
image-text-to-text
Declared base models
OpenGVLab/InternViT-300M-448px · internlm/internlm2-chat-1_8b
Formats
transformers, safetensors
Access
Open download
Publisher kind
Official
Hub parameter count
2,205,754,368safetensors.total
Repository created
2024-06-27creation is not the model's release
Repository updated
2025-03-25a commit is not a release date
Hub downloads (30 days) · likes
240,743 · 83this 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.