FitMyLLM
Liquid AI/Dense

Liquid AILFM2.5-1.2B-Thinking

Liquid AI LFM 2.5 — novel liquid foundation model with reasoning capabilities at 1.2B scale.

chatreasoningtool_useThinkingTool Use
1.2B
Parameters
122K
Context length
9
Benchmarks
6
Quantizations
33K
HF downloads
Architecture
Dense
Released
2026-01-20
Layers
16
KV Heads
8
Head Dim
64
Family
lfm

Quantization Options

Context length:
QuantBitsVRAM @ 16KQuality
Q4_K_M4.89
1.3 GB
1.2 + 0.1 KV
good
Q5_K_S5.57
1.4 GB
1.3 + 0.1 KV
good
Q5_K_M5.7
1.5 GB
1.3 + 0.1 KV
good
Q6_K6.56
1.6 GB
1.5 + 0.1 KV
excellent
Q8_08.5
1.9 GB
1.8 + 0.1 KV
lossless
FP1616
3.0 GB
2.9 + 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 LFM2.5-1.2B-Thinking now

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

Community Ratings

Loading ratings...

Benchmarks (9)

Run this model

Easiest way to get started·Beginners
Docs ↗
curl -fsSL https://ollama.com/install.sh | sh
$ollama run lfm2.5-thinking:1.2b-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

GPUs that can run this model

At Q4_K_M quantization. Sorted by minimum VRAM.

NVIDIA Tesla C870
2 GB VRAM • 76.8 GB/s
NVIDIA
NVIDIA Tesla D870
2 GB VRAM • 76.8 GB/s
NVIDIA
NVIDIA Tesla S870
2 GB VRAM • 76.8 GB/s
NVIDIA

Find the best GPU for LFM2.5-1.2B-Thinking

Build Hardware for LFM2.5-1.2B-Thinking

Liquid AI LFM 2.5 — novel liquid foundation model with reasoning capabilities at 1.2B scale.

Spec sheet

LFM2.5-1.2B-Thinking — 1.2B Dense.

Specifications
Parameters
1.2B
Architecture
Dense Transformer
Context Length
122K tokens
Capabilities
chat, reasoning, tool_use
Release Date
2026-01-20
Provider
Liquid AI
Family
lfm
VRAM requirements
QuantBPWVRAMQuality
Q4_K_M4.891.2 GB94%
Q5_K_S5.571.3 GB96%
Q5_K_M5.71.3 GB96%
Q6_K6.561.5 GB97%
Q8_08.51.8 GB100%
FP16162.9 GB100%
Benchmark scores
MMLU-PRO49.6
GPQA Diamond37.9
AIME31.7
HLE6.1
AA Intelligence8.1
AA Coding1.4
IFBench41.8
τ²-Bench19.6
SciCode4.2
Run command

Run LFM2.5-1.2B-Thinking locally with Ollama — needs 1.2 GB VRAM at Q4_K_M:

$ollama run lfm2.5-thinking:1.2b
Compatible GPUs
30 @ Q4_K_M