[whatmodel.sh](../../README.md) / LFM2.5 2.6B
# LFM2.5 2.6B
Liquid AI · released 2026-07 · 2.7B params · lfm1.0
**score 52.6** ±0.9 · #326 of 732 · from 27 benchmark results across 2 sources
Score within the [LMArena (text)](https://whatmodel.sh/?set=lmarena-text&score=set) comparison set only.
## What it's good at
● marks LFM2.5 2.6B among every model on that task or test, worst to best.
task vs the others rank counts tests
agent tasks ▁▁▃▆█●▅▅▄▄▂▁ #181 of 313 16% 6
coding ▁▁▁▃▅●█▇▅▄▃▁ #361 of 469 22% 5
chat & writing ▁▁▁▃▆▄▆█▆●▇▃ #120 of 402* 0.3% 1
math ▁▁▃▅█●▄█▄▂▁▁ #248 of 431 39% 4
web search ▅·●▅█·▃▃▃·▅▃ #12 of 15* 2.3% 1
agent tasks · #181 of 313 · 16% of its score
Beats [o1 (2024-12-17)](o1-2024-12-17.md), [o3-mini](o3-mini.md), [DeepSeek R1 0528](deepseek-r1-0528.md) here, though they rank higher overall. Trails [SmolLM3 3B](smollm3-3b.md), [Jamba 1.5 Mini](jamba-1-5-mini.md), [Jamba 1.5 Large](jamba-1-5-large.md) here, though they rank lower overall.
test result vs the others in its run difficulty counts why
BFCL V4 (December 2025 archive) 61.1% ▄·█·▄▄▄▄·●·▄ beats 7/8 medium · avg 43% 6.0% at its level · several runs
SWE-bench Verified 6.0% ▄··▄▄●··▄·▄█ beats 4/8 very hard · avg 11% 2.6% sharp · noisy
BrowseComp-ZH 9.8% ▄▄·▄●▄·█···█ beats 3/8 very hard · avg 13% 2.3% noisy
τ²-bench Telecom 90.4% ▁·▅█▆▅▂●▃▂▃▁ beats 68/89 medium · avg 35% 1.9% sharp · at its level · noisy · small run
SWE-bench Pro (split unspecified) 0.6% ●▂···▂···▄·▂ beats 3/8 very hard · avg 3.9% 1.7% sharp · too hard for it · noisy
τ³-bench Banking (maker) 5.7% ▂▂·▅●▂▃▃▃▆▂▅ beats 7/24 very hard · avg 4.0% 1.6% blunt · too hard for it
coding · #361 of 469 · 22% of its score
Trails [Code Tulu 2 13B](code-tulu-2-13b.md), [Phi-3-mini-128k-instruct (Apr 2024)](phi-3-mini-128k-instruct.md), [DeepSeek LLM 67B Chat](deepseek-llm-67b-chat.md) here, though they rank lower overall.
test result vs the others in its run difficulty counts why
LiveCodeBench v6 (window unspecif… 42.1% ▃····●·▃▅▃·▃ beats 2/8 medium · avg 38% 10% at its level · several runs
LiveCodeBench Pro 2025Q2 (Easy) 30.9% ▄···█●··▄█▄▄ beats 3/8 hard · avg 33% 4.6% at its level
LiveCodeBench Pro 2025Q2 (Medium) 0.0% ●···▂··▄▂··▂ beats 3/8 very hard · avg 2.0% 2.7% sharp · too hard for it · noisy
SWE-bench Verified 6.0% ▄··▄▄●··▄·▄█ beats 4/8 very hard · avg 11% 2.6% sharp · noisy
SWE-bench Pro (split unspecified) 0.6% ●▂···▂···▄·▂ beats 3/8 very hard · avg 3.9% 1.7% sharp · too hard for it · noisy
chat & writing · #120 of 402 · 0.3% of its score
Beats [LongCat 2.0 (1.6T-A48B)](longcat-2-0.md), [Step 3.5 Flash (196B-A11B)](step-3-5-flash.md), [GPT-5 mini](gpt-5-mini.md) here, though they rank higher overall.
test result vs the others in its run difficulty counts why
IFEval (aggregation unspecified) 93.4% ▄▄····▄··██● beats 7/8 easy · avg 76% 0.3% noisy
math · #248 of 431 · 39% of its score
Beats [GPT-4o (2024-08-06)](gpt-4o-2024-08-06.md), [Command A (03-2025)](command-a.md), [GPT-4o (2024-05-13)](gpt-4o-2024-05-13.md) here, though they rank higher overall. Trails [Gemma 7B (base)](gemma-7b-base.md), [CodeGemma 7B (v1.0)](codegemma-7b.md), [Phi-2 (2.7B)](phi-2.md) here, though they rank lower overall.
test result vs the others in its run difficulty counts why
AIME 2025 41.9% █▄●▄▄····█·▄ beats 3/8 medium · avg 42% 15% sharp · at its level · several runs
HMMT February 2026 33.7% ▅··▃●·▃··▃·█ beats 3/8 hard · avg 30% 11% sharp · at its level
AIME 2026 45.2% ▄·▄●·▄▄···█▄ beats 3/8 medium · avg 48% 8.3% sharp · at its level
MATH-500 89.6% █●▄▄▄··▄···▄ beats 3/8 very easy · avg 89% 4.4% typical
web search · #12 of 15 · 2.3% of its score
Beats [Nemotron 3 Nano 4B](nemotron-3-nano-4b.md), [Gemma 4 E4B](gemma-4-e4b.md) here, though they rank higher overall. Trails [Qwen3.5 2B](qwen3-5-2b.md), [LFM2.5 8B-A1B](lfm2-5-8b-a1b.md) here, though they rank lower overall.
test result vs the others in its run difficulty counts why
BrowseComp-ZH 9.8% ▄▄·▄●▄·█···█ beats 3/8 very hard · avg 13% 2.3% noisy
other tests (8)
test result vs the others in its run difficulty counts why
MMLU-Pro 65.2% ▂··▂●·▂▂···▂ beats 3/8 medium · avg 59% 5.1% at its level
MMLU-Redux 80.0% ▂···●··▂▂··▂ beats 5/8 easy · avg 76% 4.6% at its level
IFBench (aggregation unspecified) 59.0% ▅····▃▃·●·▃▃ beats 6/8 medium · avg 39% 4.4% at its level · several runs
OJBench 11.2% ▃····▃●··█▃▃ beats 2/8 very hard · avg 12% 3.2% typical
SuperGPQA 26.2% ●·▃·▃▅█····▃ beats 0/8 hard · avg 32% 2.9% at its level
LongBench v2 30.3% ▃······●▅▃·█ beats 1/8 very hard · avg 32% 2.5% typical
Multi-IF (aggregation unspecified) 76.8% ▄▄···▄·▄·▄█● beats 8/8 medium · avg 60% 2.1% at its level · several runs
Claw-Eval average score (maker) 62.8% ▄·▄··●▄▄▄▄▄▄ beats 3/9 medium · avg 53% 1.8% blunt · at its level
source results share of the score
openbmb/MiniCPM5-2B model card 20 81% ▇▇▇▇▇▇▇▇▇▇
LiquidAI/LFM2.5-2.6B model card (submit… 7 19% ▇▇
prior: pull toward the middle 0% ▇
weighted average = 52.6
Licences: openbmb/MiniCPM5-2B model card ([Apache-2.0](https://www.apache.org/licenses/LICENSE-2.0)) · LiquidAI/LFM2.5-2.6B model card ([LFM Open License v1.0](https://huggingface.co/LiquidAI/LFM2.5-2.6B/blob/654f9463ce32b05d0429d76fe1f580b27d4c1ac0/LICENSE))
## 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 LFM2.5 2.6B (Q4_K_M, needs ~1.67 GB, from [huggingface.co/LiquidAI](https://huggingface.co/LiquidAI/LFM2.5-2.6B-GGUF/tree/e7caca5d835a3901a8e0d63e94009429bafafdfc))
```sh
llama cli -hf LiquidAI/LFM2.5-2.6B-GGUF:Q4_K_M
```
3. Or start a local server and open the address it prints (http://localhost:9931)
```sh
llama serve -hf LiquidAI/LFM2.5-2.6B-GGUF:Q4_K_M
```
Other versions: Q8_0 (~2.87 GB) `llama cli -hf LiquidAI/LFM2.5-2.6B-GGUF:Q8_0`
### 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 LFM2.5 2.6B (Q4_K_M, needs ~1.67 GB, from [huggingface.co/LiquidAI](https://huggingface.co/LiquidAI/LFM2.5-2.6B-GGUF/tree/e7caca5d835a3901a8e0d63e94009429bafafdfc))
```sh
lms get https://huggingface.co/LiquidAI/LFM2.5-2.6B-GGUF@Q4_K_M -y
```
3. Then run lms ls to see its name, and lms chat followed by that name
Other versions: Q8_0 (~2.87 GB) `lms get https://huggingface.co/LiquidAI/LFM2.5-2.6B-GGUF@Q8_0 -y` · 4bit (~1.6 GB, Mac only) `lms get https://huggingface.co/LiquidAI/LFM2.5-2.6B-MLX-4bit -y` · 8bit (~2.9 GB, Mac only) `lms get https://huggingface.co/LiquidAI/LFM2.5-2.6B-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 LFM2.5 2.6B (Q4_K_M, needs ~1.67 GB, from [huggingface.co/LiquidAI](https://huggingface.co/LiquidAI/LFM2.5-2.6B-GGUF/tree/e7caca5d835a3901a8e0d63e94009429bafafdfc))
```sh
ollama run hf.co/LiquidAI/LFM2.5-2.6B-GGUF:Q4_K_M
```
Other versions: Q8_0 (~2.87 GB) `ollama run hf.co/LiquidAI/LFM2.5-2.6B-GGUF:Q8_0`
### 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 LFM2.5 2.6B (4bit, needs ~1.6 GB, Mac only, from [huggingface.co/LiquidAI](https://huggingface.co/LiquidAI/LFM2.5-2.6B-MLX-4bit/tree/04efa23776ce61ec34ec95ec34c859854c89542b))
```sh
mlx_lm.chat --model LiquidAI/LFM2.5-2.6B-MLX-4bit
```
3. Or start an OpenAI-compatible server at http://localhost:8080/v1
```sh
mlx_lm.server --model LiquidAI/LFM2.5-2.6B-MLX-4bit
```
Other versions: 8bit (~2.9 GB, Mac only) `mlx_lm.chat --model LiquidAI/LFM2.5-2.6B-MLX-8bit`
Model repository and documentation: [huggingface.co/LiquidAI](https://huggingface.co/LiquidAI/LFM2.5-2.6B)
[interactive version](https://whatmodel.sh/model/lfm2-5-2-6b) · [JSON](https://whatmodel.sh/api?format=json&all=1)