whatmodel.sh / LFM2 350M

LFM2 350M

+ contribute
Update model informationAdd a place to run itAdd a test resultPropose a benchmarkCorrect a benchmark scoreEdit data on GitHub

LFM2 350M is an open-weights language model from Liquid AI, released July 2025. Its published size is 0.354B parameters and a 33k-token context window. Licence: lfm1.0.

Catalogue identity notes

LiquidAI/LFM2-350M @f37d3f5 (created 2025-07-10, licence lfm1.0). Liquid AI LFM2 hybrid (conv + attention) instruct model, July-Sept 2025; earlier generation than the catalogued LFM2.5 models. Lab rows from the LFM2-1.2B card's own LFM2 columns (3-model runs) plus independent EuroEval rows.

Model information sources

Release date ↗ · Weights licence ↗

HF safetensors metadata 354,483,968 parameters (LiquidAI/LFM2-350M @f37d3f5).

30.4±1.7#640 of 719 LLMsfrom 10 benchmark results across 3 sources

$ how it compares

The LLMs ranked either side of LFM2 350M, out of 828 with a score, then more from Liquid AI.

#modelscorewhereRAMreleased
735Llama 2 7B (base) Meta31±1local~5.5GB2023-07
736OPT-66B Meta31±1local~141GB2022-05
737LFM2 350M Liquid AI30±2local~1.7GB2025-07
738SmolLM2 1.7B Instruct Hugging Face30±1local~2.6GB2024-10
739Llama 3.2 1B Instruct Meta30±1api $0.020/M +local~2.2GB2024-09
more from Liquid AI
370LFM2.5 2.6B Liquid AI53±1local~3.2GB2026-07
375LFM2.5 8B-A1B Liquid AI52±1local~6.5GB2026-05
398LFM2 24B-A2B Liquid AI51±1local~16GB2026-02

see all LLMs · find the ones that fit your machine

$ what it's good at

chat & writing▁▁▁▃▅▄▅█▆▆▇▃ #295 of 393 * · 10% of its score · 1 test

Beats Ministral 3 14B Reasoning 2512, Gemma 4 E4B, OpenThinker3 7B here, though they rank higher overall.

testresultvs the others in its rundifficultycountswhy
IFEval (aggregation unspecified) lab65.1% ✎▄····█▄···▄█ beats 3/6easy · avg 82%▇▇▇▇ 10%at its level
math▁▂▃▅▆▅▄█▄▁▁▁ #371 of 398 * · 36% of its score · 2 tests

Beats Command R (03-2024), Molmo 7B-D (0924), Mistral 7B Instruct v0.3 here, though they rank higher overall. Trails SmolLM2 1.7B Instruct here, though they rank lower overall.

testresultvs the others in its rundifficultycountswhy
MGSM lab29.5% ✎█···█▄··▄··▄ beats 1/6easy · avg 77%▇▇▇▇▇▇▇▇▇▇ 24%at its level
GSM8K lab30.1% ✎▄·█···▄·▄··█ beats 0/6easy · avg 83%▇▇▇▇▇ 12%sharp · at its level
other tests7 not in a task · 53% of its score
testresultvs the others in its rundifficultycountswhy
MMLU lab43.4% ✎█·███·█··█·█ beats 1/6easy · avg 75%▇▇▇▇▇▇▇▇ 20%typical
MMMLU lab38.0% ✎▄···▄·▄▄··▄█ beats 2/6medium · avg 59%▇▇▇▇ 9.4%typical
IFBench (aggregation unspecified) lab16.4% ✎▄▄·▄····▄·█▄ beats 0/6medium · avg 40%▇▇▇▇ 8.8%typical
GPQA (subset unspecified) lab27.5% ✎█······▄▄█·▄ beats 2/6hard · avg 39%▇▇ 4.1%sharp · too hard for it · noisy
Open PT-LLM Leaderboard average (9 Portuguese tasks)32.2% ✎▃▁▁▁▁▁▁▁▂█▅▁ beats 18/276medium · avg 73%▇▇ 4.1%noisy
EuroEval Life in the UK (English knowledge MC), test split, 5-shot, accuracy37.8% ✎▁▁▁▁▁▃▁▂▄█▄▃ beats 13/146easy · avg 83%▇▇ 3.8%at its level · noisy
EuroEval HellaSwag (English), test split, 5-shot generative MC, accuracy25.6% ✎▃·▁▁▂▁▂▅█▇▄▁ beats 7/100medium · avg 68%▇ 2.3%noisy

$ where to run it

llama.cpp · recommended

  1. Install llama.cpp · Homebrew and winget installs name the commands llama-cli and llama-server instead.
    macOS / Linux
    curl -LsSf https://llama.app/install.sh | sh
    Windows
    irm https://llama.app/install.ps1 | iex

    other ways to install ↗

  2. Run LFM2 350M · needs ~0.23 GB · from huggingface.co/LiquidAI (official)
    chat in the terminal
    llama cli -hf LiquidAI/LFM2-350M-GGUF:Q4_K_M
  3. Or start a local server and open the address it prints (http://localhost:9931)
    llama serve -hf LiquidAI/LFM2-350M-GGUF:Q4_K_M

LM Studio

version

  1. Install LM Studio · Open the app once before using the lms command.

    download it (macOS / Windows / Linux) ↗

  2. Run LFM2 350M · needs ~0.23 GB · from huggingface.co/LiquidAI (official)

    open in LM Studio ↗

  3. Or download it in the terminal
    lms get https://huggingface.co/LiquidAI/LFM2-350M-GGUF@Q4_K_M -y
  4. Then run lms ls to see its name, and lms chat followed by that name

Ollama

  1. Install Ollama

    download it (macOS / Windows) ↗

    Linux
    curl -fsSL https://ollama.com/install.sh | sh
  2. Run LFM2 350M · needs ~0.23 GB · from huggingface.co/LiquidAI (official)
    chat in the terminal
    ollama run hf.co/LiquidAI/LFM2-350M-GGUF:Q4_K_M

MLX (Mac)

version

  1. Install MLX (Mac) · Apple silicon Macs only, text only.
    with uv
    uv tool install mlx-lm
    or
    pip install mlx-lm
  2. Run LFM2 350M · needs ~0.2 GB, Mac only · from huggingface.co/mlx-community
    chat in the terminal
    mlx_lm.chat --model mlx-community/LFM2-350M-4bit
  3. Or start an OpenAI-compatible server at http://localhost:8080/v1
    mlx_lm.server --model mlx-community/LFM2-350M-4bit

Official weights and documentation: huggingface.co/LiquidAI ↗

$ how the score was worked out · LiquidAI/LFM2-1.2B model card 89%, EuroEval leaderboards (result… 6%, Open Portuguese LLM Leaderboa… 4%

Each result implies a score on its own, after adjusting for how hard its test is and who ran it. The score is their weighted average, with a small pull toward the middle for thinly tested models. Each test's weight is its counts column above; here it's added up by who ran the tests. How it works

sourceresultsshare of the scorelicence
LiquidAI/LFM2-1.2B model card · lab-reported7▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇ 89%LFM Open License v1.0
EuroEval leaderboards (results collection), English generative leaderboard2▇▇ 6%MIT
Open Portuguese LLM Leaderboard (archived)1▇ 4%Apache-2.0
prior · pull toward the middle▇ 1%
weighted average= 30.4

$ improve this page

Update model informationAdd a place to run itAdd a test resultPropose a benchmarkCorrect a benchmark scoreEdit data on GitHub

Use a short GitHub form or edit the data through a pull request. Every number needs a source; here's which ones we can use.

data updated · sources and licences · json