FitMyLLM
Liquid AI/Dense

Liquid AILFM3-VL 2B

LFM3-VL 2B — Liquid AI's LFM3 vision-language model, the first LFM vision model to take video as well as images, with a thinking mode for charts, maths and spatial questions. Built on LFM3 1.7B with a new in-house vision encoder.

chatvisionreasoningmultilingual
1.92B
Parameters
125K
Context length
3
Benchmarks
6
Quantizations
11K
HF downloads
Architecture
Dense
Released
2026-10-27
Layers
24
KV Heads
4
Head Dim
128
Family
lfm

Quantization Options

Context length:
QuantBitsVRAM @ 16KQuality
Q4_K_M4.89
1.8 GB
1.7 + 0.1 KV
good
Q5_K_S5.57
2.0 GB
1.8 + 0.1 KV
good
Q5_K_M5.7
2.0 GB
1.9 + 0.1 KV
good
Q6_K6.56
2.2 GB
2.1 + 0.1 KV
excellent
Q8_08.5
2.7 GB
2.5 + 0.1 KV
lossless
FP1616
4.5 GB
4.3 + 0.1 KV
lossless

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

Deploying for a team or in production? Size GPUs, cost & scaling in Enterprise →
▸ READY TO RUN THIS?RENT BY THE HOUR

RENT A GPU AND RUN LFM3-VL 2B NOW

Rent on Vast.ai →Or RunPod →

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

Community Ratings

Loading ratings...

Benchmarks (3)

MMBench86.1
MMMU55.7
MMMU Pro45.7

Run this model

▸Easiest way to get started·Beginners
DOCS ↗
curl -fsSL https://ollama.com/install.sh | sh
$ollama run lfm3-vl:2b-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
HuggingFaceOllama LibraryGGUF DownloadsBuild Hardware

GPUs that can run this model

At Q4_K_M quantization. Sorted by minimum VRAM.

NVIDIA GeForce GTX 260 OEM
2 GB VRAM • 112.9 GB/s
NVIDIA
AmazonNeweggeBayPCPartPickerRent on RunPodRent on Vast.ai
NVIDIA Quadro 4000 Mac Edition
2 GB VRAM • 89.9 GB/s
NVIDIA
AmazonNeweggeBayPCPartPickerRent on RunPodRent on Vast.ai
AMD FireGL V8650
2 GB VRAM • 111 GB/s
AMD
AmazonNeweggeBayPCPartPickerRent on RunPodRent on Vast.ai
NVIDIA GeForce GTX 285 X2
2 GB VRAM • 148 GB/s
NVIDIA
AmazonNeweggeBayPCPartPickerRent on RunPodRent on Vast.ai
NVIDIA GeForce GTX 480M
2 GB VRAM • 77 GB/s
NVIDIA
AmazonNeweggeBayPCPartPickerRent on RunPodRent on Vast.ai
NVIDIA Quadro 5000M
2 GB VRAM • 77 GB/s
NVIDIA
AmazonNeweggeBayPCPartPickerRent on RunPodRent on Vast.ai
AMD Radeon HD 6950
2 GB VRAM • 160 GB/s
AMD
AmazonNeweggeBayPCPartPickerRent on RunPodRent on Vast.ai
AMD Radeon HD 6970
2 GB VRAM • 176 GB/s
AMD
AmazonNeweggeBayPCPartPickerRent on RunPodRent on Vast.ai
Intel Aubrey Isle
2 GB VRAM • 154 GB/s
INTEL
AmazonNeweggeBayPCPartPickerRent on RunPodRent on Vast.ai
AMD Radeon HD 5870 Eyefinity 6
2 GB VRAM • 154 GB/s
AMD
AmazonNeweggeBayPCPartPickerRent on RunPodRent on Vast.ai
NVIDIA GeForce GTX 485M
2 GB VRAM • 96 GB/s
NVIDIA
AmazonNeweggeBayPCPartPickerRent on RunPodRent on Vast.ai
NVIDIA GeForce GTX 580M
2 GB VRAM • 96 GB/s
NVIDIA
AmazonNeweggeBayPCPartPickerRent on RunPodRent on Vast.ai
NVIDIA Quadro 1000M
2 GB VRAM • 29 GB/s
NVIDIA
AmazonNeweggeBayPCPartPickerRent on RunPodRent on Vast.ai
NVIDIA Quadro 2000M
2 GB VRAM • 29 GB/s
NVIDIA
AmazonNeweggeBayPCPartPickerRent on RunPodRent on Vast.ai
NVIDIA Quadro 3000M
2 GB VRAM • 80 GB/s
NVIDIA
AmazonNeweggeBayPCPartPickerRent on RunPodRent on Vast.ai
NVIDIA Quadro 3000M X2
2 GB VRAM • 80 GB/s
NVIDIA
AmazonNeweggeBayPCPartPickerRent on RunPodRent on Vast.ai
NVIDIA Quadro 4000M
2 GB VRAM • 80 GB/s
NVIDIA
AmazonNeweggeBayPCPartPickerRent on RunPodRent on Vast.ai
AMD Radeon HD 6550A
2 GB VRAM • 26 GB/s
AMD
AmazonNeweggeBayPCPartPickerRent on RunPodRent on Vast.ai
AMD Radeon HD 6570
2 GB VRAM • 64 GB/s
AMD
AmazonNeweggeBayPCPartPickerRent on RunPodRent on Vast.ai
AMD Radeon HD 6850 X2
2 GB VRAM • 134 GB/s
AMD
AmazonNeweggeBayPCPartPickerRent on RunPodRent on Vast.ai
AMD Radeon HD 6970M Mac Edition
2 GB VRAM • 115 GB/s
AMD
AmazonNeweggeBayPCPartPickerRent on RunPodRent on Vast.ai
AMD Radeon HD 6970M X2
2 GB VRAM • 115 GB/s
AMD
AmazonNeweggeBayPCPartPickerRent on RunPodRent on Vast.ai
AMD Radeon HD 6990
2 GB VRAM • 160 GB/s
AMD
AmazonNeweggeBayPCPartPickerRent on RunPodRent on Vast.ai
AMD Radeon HD 6990M
2 GB VRAM • 115 GB/s
AMD
AmazonNeweggeBayPCPartPickerRent on RunPodRent on Vast.ai
NVIDIA GeForce GTX 660
2 GB VRAM • 144 GB/s
NVIDIA
AmazonNeweggeBayPCPartPickerRent on RunPodRent on Vast.ai
NVIDIA GeForce GTX 660 OEM
2 GB VRAM • 179 GB/s
NVIDIA
AmazonNeweggeBayPCPartPickerRent on RunPodRent on Vast.ai
NVIDIA GeForce GTX 660 Ti
2 GB VRAM • 144 GB/s
NVIDIA
AmazonNeweggeBayPCPartPickerRent on RunPodRent on Vast.ai
NVIDIA GeForce GTX 660M
2 GB VRAM • 80 GB/s
NVIDIA
AmazonNeweggeBayPCPartPickerRent on RunPodRent on Vast.ai
NVIDIA GeForce GTX 670
2 GB VRAM • 192 GB/s
NVIDIA
AmazonNeweggeBayPCPartPickerRent on RunPodRent on Vast.ai
NVIDIA GeForce GTX 675M
2 GB VRAM • 96 GB/s
NVIDIA
AmazonNeweggeBayPCPartPickerRent on RunPodRent on Vast.ai

Find the best GPU for LFM3-VL 2B

Build Hardware for LFM3-VL 2B
▸ SPEC SHEET

LFM3-VL 2B — 1.92B Dense.

▸ SPECIFICATIONS
PARAMETERS
1.92B
ARCHITECTURE
Dense Transformer
CONTEXT LENGTH
125K tokens
CAPABILITIES
chat, vision, reasoning, multilingual
RELEASE DATE
2026-10-27
PROVIDER
Liquid AI
FAMILY
lfm
▸ VRAM REQUIREMENTS
QUANTBPWVRAMQUALITY
Q4_K_M4.891.7 GB94%
Q5_K_S5.571.8 GB96%
Q5_K_M5.71.9 GB96%
Q6_K6.562.1 GB97%
Q8_08.52.5 GB100%
FP16164.3 GB100%
§ 01BENCHMARK SCORES
MMMU55.7
MMBench86.1
mmmu_pro45.7
§ 02RUN COMMAND

Run LFM3-VL 2B locally with Ollama — needs 1.7 GB VRAM at Q4_K_M:

$ollama run lfm3-vl:2b
§ 03COMPATIBLE GPUs
30 @ Q4_K_M
NVIDIA GeForce GTX 260 OEM
2 GB · 112.9 GB/s
NVIDIA Quadro 4000 Mac Edition
2 GB · 89.9 GB/s
AMD FireGL V8650
2 GB · 111 GB/s
NVIDIA GeForce GTX 285 X2
2 GB · 148 GB/s
NVIDIA GeForce GTX 480M
2 GB · 77 GB/s
NVIDIA Quadro 5000M
2 GB · 77 GB/s
AMD Radeon HD 6950
2 GB · 160 GB/s
AMD Radeon HD 6970
2 GB · 176 GB/s
Intel Aubrey Isle
2 GB · 154 GB/s
AMD Radeon HD 5870 Eyefinity 6
2 GB · 154 GB/s
NVIDIA GeForce GTX 485M
2 GB · 96 GB/s
NVIDIA GeForce GTX 580M
2 GB · 96 GB/s
NVIDIA Quadro 1000M
2 GB · 29 GB/s
NVIDIA Quadro 2000M
2 GB · 29 GB/s
NVIDIA Quadro 3000M
2 GB · 80 GB/s