LiquidAI/ lfm2.5-350m

533 Downloads Updated 4 weeks ago

Ultra-fast edge model for data extraction and tool use

ollama run LiquidAI/lfm2.5-350m

curl http://localhost:11434/api/chat \
  -d '{
    "model": "LiquidAI/lfm2.5-350m",
    "messages": [{"role": "user", "content": "Hello!"}]
  }'

from ollama import chat

response = chat(
    model='LiquidAI/lfm2.5-350m',
    messages=[{'role': 'user', 'content': 'Hello!'}],
)
print(response.message.content)

import ollama from 'ollama'

const response = await ollama.chat({
  model: 'LiquidAI/lfm2.5-350m',
  messages: [{role: 'user', content: 'Hello!'}],
})
console.log(response.message.content)

Models

Name

6 models

Size / Usage

Context

Input

lfm2.5-350m:latest

379MB · 125K context window · Text · 4 weeks ago

lfm2.5-350m:latest

379MB

125K

Text

lfm2.5-350m:q4_0

219MB · 125K context window · Text · 4 weeks ago

lfm2.5-350m:q4_0

219MB

125K

Text

lfm2.5-350m:q4_k_m

229MB · 125K context window · Text · 4 weeks ago

lfm2.5-350m:q4_k_m

229MB

125K

Text

lfm2.5-350m:q5_k_m

260MB · 125K context window · Text · 4 weeks ago

lfm2.5-350m:q5_k_m

260MB

125K

Text

lfm2.5-350m:q6_k

293MB · 125K context window · Text · 4 weeks ago

lfm2.5-350m:q6_k

293MB

125K

Text

lfm2.5-350m:q8_0

379MB · 125K context window · Text · 4 weeks ago

lfm2.5-350m:q8_0

379MB

125K

Text

Readme

LFM2.5-350M

LFM2.5 is a new family of hybrid models designed for on-device deployment. It builds on the LFM2 architecture with extended pre-training and reinforcement learning.

Best-in-class performance: A 350M model rivaling much larger models, bringing high-quality AI to your pocket.
Fast edge inference: 313 tok/s decode on AMD CPU, 188 tok/s on Snapdragon Gen4. Runs under 1GB of memory with day-one support for llama.cpp, MLX, and vLLM.
Scaled training: Extended pre-training from 10T to 28T tokens and large-scale multi-stage reinforcement learning.

Find more information about LFM2.5-350M in our blog post.

🗒️ Model Details

LFM2.5-350M is a general-purpose text-only model with the following features:

Number of parameters: 350M
Number of layers: 16 (10 double-gated LIV convolution blocks + 6 GQA blocks)
Training budget: 28T tokens
Context length: 32,768 tokens
Vocabulary size: 65,536
Knowledge cutoff: Mid-2024
Languages: English, Arabic, Chinese, French, German, Japanese, Korean, Portuguese, Spanish
Generation parameters:
- temperature: 0.1
- top_k: 50
- repetition_penalty: 1.05

We recommend using it for data extraction, structured outputs, and tool use. It is not recommended for knowledge-intensive tasks and programming.

📊 Performance

Benchmarks

Model	GPQA Diamond	MMLU-Pro	IFEval	IFBench	Multi-IF
LFM2.5-350M	30.64	20.01	76.96	40.69	44.92
LFM2-350M	27.58	19.29	64.96	18.20	32.92
Granite 4.0-H-350M	22.32	13.14	61.27	17.22	28.70
Granite 4.0-350M	25.91	12.84	53.48	15.98	24.21
Qwen3.5-0.8B (Instruct)	27.41	37.42	59.94	22.87	41.68
Qwen3.5-0.8B (Thinking)	19.29	-*	32.93	22.00	26.44
Gemma 3 1B IT	23.89	14.04	63.49	20.33	44.25

Model	CaseReportBench	BFCLv3	BFCLv4	τ²-Bench Telecom	τ²-Bench Retail
LFM2.5-350M	32.45	44.11	21.86	18.86	17.84
LFM2-350M	11.67	22.95	12.29	10.82	5.56
Granite 4.0-H-350M	12.44	43.07	13.28	13.74	6.14
Granite 4.0-350M	0.84	39.58	13.73	2.92	6.14
Qwen3.5-0.8B (Instruct)	13.83	35.08	18.70	12.57	6.14
Qwen3.5-0.8B (Thinking)	0.39	39.64	25.39	14.33	7.02
Gemma 3 1B IT	2.28	16.61	7.17	9.36	6.43

*Evaluation could not be completed due to doom looping.

CPU Inference

GPU Inference

📬 Contact

Got questions or want to connect? Join our Discord community
If you are interested in custom solutions with edge deployment, please contact our sales team.

<div align="center">
  <img 
    src="https://cdn-uploads.huggingface.co/production/uploads/61b8e2ba285851687028d395/2b08LKpev0DNEk6DlnWkY.png" 
    alt="Liquid AI" 
    style="width: 100%; max-width: 100%; height: auto; display: inline-block; margin-bottom: 0.5em; margin-top: 0.5em;"
  />
</div>

# LFM2.5-350M

LFM2.5 is a new family of hybrid models designed for **on-device deployment**. It builds on the LFM2 architecture with extended pre-training and reinforcement learning.

- **Best-in-class performance**: A 350M model rivaling much larger models, bringing high-quality AI to your pocket.
- **Fast edge inference**: 313 tok/s decode on AMD CPU, 188 tok/s on Snapdragon Gen4. Runs under 1GB of memory with day-one support for llama.cpp, MLX, and vLLM.
- **Scaled training**: Extended pre-training from 10T to 28T tokens and large-scale multi-stage reinforcement learning.

Find more information about LFM2.5-350M in our [blog post](https://www.liquid.ai/blog/lfm2-5-350m-no-size-left-behind).

![](https://cdn-uploads.huggingface.co/production/uploads/61b8e2ba285851687028d395/mx39JYUuCa1ehaucRFT7d.png)

## 🗒️ Model Details

LFM2.5-350M is a general-purpose text-only model with the following features:

- **Number of parameters**: 350M
- **Number of layers**: 16 (10 double-gated LIV convolution blocks + 6 GQA blocks)
- **Training budget**: 28T tokens
- **Context length**: 32,768 tokens
- **Vocabulary size**: 65,536
- **Knowledge cutoff**: Mid-2024
- **Languages**: English, Arabic, Chinese, French, German, Japanese, Korean, Portuguese, Spanish
- **Generation parameters**:
  - `temperature: 0.1`
  - `top_k: 50`
  - `repetition_penalty: 1.05`

We recommend using it for data extraction, structured outputs, and tool use. It is not recommended for knowledge-intensive tasks and programming.

## 📊 Performance

### Benchmarks

| Model | GPQA Diamond | MMLU-Pro | IFEval | IFBench | Multi-IF |
|---|---|---|---|---|---|
| LFM2.5-350M | 30.64 | 20.01 | 76.96 | 40.69 | 44.92 |
| LFM2-350M | 27.58 | 19.29 | 64.96 | 18.20 | 32.92 |
| Granite 4.0-H-350M | 22.32 | 13.14 | 61.27 | 17.22 | 28.70 |
| Granite 4.0-350M | 25.91 | 12.84 | 53.48 | 15.98 | 24.21 |
| Qwen3.5-0.8B (Instruct) | 27.41 | 37.42 | 59.94 | 22.87 | 41.68 |
| Qwen3.5-0.8B (Thinking) | 19.29 | -* | 32.93 | 22.00 | 26.44 |
| Gemma 3 1B IT | 23.89 | 14.04 | 63.49 | 20.33 | 44.25 |

| Model | CaseReportBench | BFCLv3 | BFCLv4 | τ²-Bench Telecom | τ²-Bench Retail |
|---|---|---|---|---|---|
| LFM2.5-350M | 32.45 | 44.11 | 21.86 | 18.86 | 17.84 |
| LFM2-350M | 11.67 | 22.95 | 12.29 | 10.82 | 5.56 |
| Granite 4.0-H-350M | 12.44 | 43.07 | 13.28 | 13.74 | 6.14 |
| Granite 4.0-350M | 0.84 | 39.58 | 13.73 | 2.92 | 6.14 |
| Qwen3.5-0.8B (Instruct) | 13.83 | 35.08 | 18.70 | 12.57 | 6.14 |
| Qwen3.5-0.8B (Thinking) | 0.39 | 39.64 | 25.39 | 14.33 | 7.02 |
| Gemma 3 1B IT | 2.28 | 16.61 | 7.17 | 9.36 | 6.43 |

<i>*Evaluation could not be completed due to doom looping.</i>

### CPU Inference

![](https://cdn-uploads.huggingface.co/production/uploads/61b8e2ba285851687028d395/tlt5UmogSZjbMGC6YEYuO.png)

### GPU Inference

![](https://cdn-uploads.huggingface.co/production/uploads/61b8e2ba285851687028d395/1vzwlXxvFr8lZWmu5jDSx.png)

## 📬 Contact

- Got questions or want to connect? [Join our Discord community](https://discord.com/invite/liquid-ai)
- If you are interested in custom solutions with edge deployment, please contact [our sales team](https://www.liquid.ai/contact).

Paste, drop or click to upload images (.png, .jpeg, .jpg, .svg, .gif)