vikhyatk/Dense

VMoondream2 1.9B

Moondream2 — tiny vision model. Runs on any GPU, surprisingly good at image description.

visionchatThinking
1.9B
Parameters
2K
Context length
8
Benchmarks
6
Quantizations
890K
HF downloads
Architecture
Dense
Released
2024-03-06
Layers
24
KV Heads
4
Head Dim
64
Family
other

Quantization Options

QuantBitsVRAMQuality
Q4_K_M4.891.6 GBgood
Q5_K_S5.571.8 GBgood
Q5_K_M5.71.8 GBgood
Q6_K6.562.0 GBexcellent
Q8_08.52.5 GBlossless
FP16164.3 GBlossless

Select your GPU above to see speed estimates and compatibility for each quantization.

READY TO RUN THIS?RENT BY THE HOUR

RENT A GPU AND RUN MOONDREAM2 1.9B NOW

Spin up an A100 / H100 / 4090 in ~60s. Pay by the second. Cancel anytime.

Community Ratings

Loading ratings...

Benchmarks (8)

MMBench58.2
IFEval35.0
MMMU32.0
BBH28.0
MMLU-PRO25.0
GPQA20.0
MUSR8.0
MATH3.0

Run this model

Easiest way to get started·Beginners
DOCS ↗
curl -fsSL https://ollama.com/install.sh | sh
$ollama run moondream:1.8b-v2-q4_K_M

Downloads and runs automatically. Add --verbose for speed stats.

▸ SETUP GUIDE
>_

Auto-setup with fitmyllm CLI

Detects your GPU, recommends the best model, downloads it, and starts chatting — zero config. Benchmarks your speed and contributes anonymous data to improve predictions.

pip install fitmyllmthen run fitmyllmLearn more
Auto-detect GPULive tok/s in chatSpeed benchmarks9 inference engines

GPUs that can run this model

At Q4_K_M quantization. Sorted by minimum VRAM.

Find the best GPU for Moondream2 1.9B

Build Hardware for Moondream2 1.9B

Moondream2 — tiny vision model. Runs on any GPU, surprisingly good at image description.

▸ SPEC SHEET

Moondream2 1.9B1.9B Dense.

▸ SPECIFICATIONS
PARAMETERS
1.9B
ARCHITECTURE
Dense Transformer
CONTEXT LENGTH
2K tokens
CAPABILITIES
vision, chat
RELEASE DATE
2024-03-06
PROVIDER
vikhyatk
FAMILY
other
▸ VRAM REQUIREMENTS
QUANTBPWVRAMQUALITY
Q4_K_M4.891.6 GB94%
Q5_K_S5.571.8 GB96%
Q5_K_M5.71.8 GB96%
Q6_K6.562.0 GB97%
Q8_08.52.5 GB100%
FP16164.3 GB100%
§ 01BENCHMARK SCORES
MMLU-PRO25.0
MATH3.0
IFEval35.0
BBH28.0
MMMU32.0
GPQA20.0
MUSR8.0
MMBench58.2
§ 02RUN COMMAND

Run Moondream2 1.9B locally with Ollama — needs 1.6 GB VRAM at Q4_K_M:

$ollama run moondream:1.8b