[whatmodel.sh](../../README.md) / LFM2 700M
# LFM2 700M
Liquid AI · released 2025-07 · 0.742B params · lfm1.0
**score 37.7** ±1.3 · #601 of 732 · from 10 benchmark results across 3 sources
Score within the [LMArena (text)](https://whatmodel.sh/?set=lmarena-text&score=set) comparison set only.
## What it's good at
● marks LFM2 700M among every model on that task or test, worst to best.
task vs the others rank counts tests
chat & writing ▁▁▁▃▆▄●█▆▇▇▃ #284 of 402* 6.4% 1
math ▁▁▃●█▆▄█▄▂▁▁ #336 of 431* 41% 2
chat & writing · #284 of 402 · 6.4% of its score
Beats [Ministral 3 14B Reasoning 2512](ministral-3-14b-reasoning.md), [Ministral 3 14B](ministral-3-14b.md), [Ministral 3 8B](ministral-3-8b.md) here, though they rank higher overall. Trails [EXAONE 3.5 2.4B Instruct](exaone-3-5-2-4b.md) here, though they rank lower overall.
test result vs the others in its run difficulty counts why
IFEval (aggregation unspecified) … 72.2% ▄····█▄···●█ beats 4/6 easy · avg 82% 6.4% at its level
math · #336 of 431 · 41% of its score
Beats [GPT-4 (0613)](gpt-4-0613.md), [Athene-70B](athene-70b.md), [Llama 3 70B Instruct](llama-3-70b.md) here, though they rank higher overall. Trails [Gemma 2B (base)](gemma-2b-base.md), [Stable Code 3B](stable-code-3b.md), [Zephyr 7B beta](zephyr-7b-beta.md) here, though they rank lower overall.
test result vs the others in its run difficulty counts why
MGSM (maker) 45.4% █···█●··▄··▄ beats 4/6 easy · avg 83% 30% sharp · at its level
GSM8K (maker) 46.4% ▄·█···●·▄··█ beats 3/6 easy · avg 83% 11% sharp · at its level
other tests (7)
test result vs the others in its run difficulty counts why
MMLU (maker) 49.9% █·███·●··█·█ beats 4/6 easy · avg 74% 20% at its level
IFBench (aggregation unspecified)… 20.6% ▄▄·▄····▄·●▄ beats 4/6 medium · avg 38% 8.3% typical
MMMLU (maker) 43.3% ▄···▄·▄▄··●█ beats 4/6 medium · avg 57% 8.3% typical
GPQA (subset unspecified) (maker) 28.5% █······▄▄●·▄ beats 4/6 hard · avg 37% 5.2% sharp · too hard for it · noisy
Open PT-LLM Leaderboard average (… 55.0% ▃▁▁▁▁▁▁▁●█▅▁ beats 76/276 medium · avg 72% 4.0% at its level · noisy
EuroEval Life in the UK (English … 42.1% ▁▁▁▁▁●▁▂▄█▄▃ beats 14/146 easy · avg 82% 3.1% at its level · noisy
EuroEval HellaSwag (English), tes… 25.1% ●·▁▁▂▁▂▅█▇▄▁ beats 5/100 medium · avg 67% 2.7% sharp · noisy
source results share of the score
LiquidAI/LFM2-1.2B model card (submitte… 7 89% ▇▇▇▇▇▇▇▇▇▇▇
EuroEval leaderboards 2 6% ▇
Open Portuguese LLM Leaderboard 1 4% ▇
prior: pull toward the middle 1% ▇
weighted average = 37.7
Licences: LiquidAI/LFM2-1.2B model card ([LFM Open License v1.0](https://huggingface.co/LiquidAI/LFM2-1.2B/blob/40f3da0d0164913923aee9462c23077868b816a3/LICENSE)) · EuroEval leaderboards ([MIT](https://github.com/EuroEval/leaderboards/blob/1a77f1d2f8ba1ba05ab73287fb090d76bf41b2e0/LICENSE)) · Open Portuguese LLM Leaderboard ([Apache-2.0](../../licenses/hf-leaderboard-spaces-Apache-2.0.txt))
## How to run it
### llama.cpp (recommended)
1. Install llama.cpp. Homebrew and winget installs name the commands llama-cli and llama-server instead. [Other ways to install](https://github.com/ggml-org/llama.cpp/blob/master/docs/install.md)
```sh
# macOS / Linux
curl -LsSf https://llama.app/install.sh | sh
# Windows
irm https://llama.app/install.ps1 | iex
```
2. Run LFM2 700M (Q4_K_M, needs ~0.47 GB, from [huggingface.co/LiquidAI](https://huggingface.co/LiquidAI/LFM2-700M-GGUF/tree/fd39e80d7a5ac61494ffff577e61bbbfddbd0d02))
```sh
llama cli -hf LiquidAI/LFM2-700M-GGUF:Q4_K_M
```
3. Or start a local server and open the address it prints (http://localhost:9931)
```sh
llama serve -hf LiquidAI/LFM2-700M-GGUF:Q4_K_M
```
### LM Studio
1. Install LM Studio. Open the app once before using the lms command. [Download it (macOS / Windows / Linux)](https://lmstudio.ai/download)
2. Run LFM2 700M (Q4_K_M, needs ~0.47 GB, from [huggingface.co/LiquidAI](https://huggingface.co/LiquidAI/LFM2-700M-GGUF/tree/fd39e80d7a5ac61494ffff577e61bbbfddbd0d02))
```sh
lms get https://huggingface.co/LiquidAI/LFM2-700M-GGUF@Q4_K_M -y
```
3. Then run lms ls to see its name, and lms chat followed by that name
Other versions: 4bit (~0.4 GB, Mac only) `lms get https://huggingface.co/mlx-community/LFM2-700M-4bit -y` · 8bit (~0.8 GB, Mac only) `lms get https://huggingface.co/lmstudio-community/LFM2-700M-MLX-8bit -y`
### Ollama
1. Install Ollama. [Download it (macOS / Windows)](https://ollama.com/download)
```sh
# Linux
curl -fsSL https://ollama.com/install.sh | sh
```
2. Run LFM2 700M (Q4_K_M, needs ~0.47 GB, from [huggingface.co/LiquidAI](https://huggingface.co/LiquidAI/LFM2-700M-GGUF/tree/fd39e80d7a5ac61494ffff577e61bbbfddbd0d02))
```sh
ollama run hf.co/LiquidAI/LFM2-700M-GGUF:Q4_K_M
```
### MLX (Mac)
1. Install MLX (Mac). Apple silicon Macs only, text only.
```sh
# with uv
uv tool install mlx-lm
# or
pip install mlx-lm
```
2. Run LFM2 700M (4bit, needs ~0.4 GB, Mac only, from [huggingface.co/mlx-community](https://huggingface.co/mlx-community/LFM2-700M-4bit/tree/6b32895938db95f2d40f2d0d70c10117fba5297a))
```sh
mlx_lm.chat --model mlx-community/LFM2-700M-4bit
```
3. Or start an OpenAI-compatible server at http://localhost:8080/v1
```sh
mlx_lm.server --model mlx-community/LFM2-700M-4bit
```
Other versions: 8bit (~0.8 GB, Mac only) `mlx_lm.chat --model lmstudio-community/LFM2-700M-MLX-8bit`
Model repository and documentation: [huggingface.co/LiquidAI](https://huggingface.co/LiquidAI/LFM2-700M)
[interactive version](https://whatmodel.sh/model/lfm2-700m) · [JSON](https://whatmodel.sh/api?format=json&all=1)