# LFM 2.5 24B A2B

> Liquid AI LFM2.5-24B-A2B is an open-weight reasoning mixture-of-experts model with 24B total and about 2B active parameters per token. It always reasons before it answers and is built for fast agents, tool use and structured output, with a 128K context window.

**ID**: `liquid/lfm-2.5-24b-a2b`  
**Creator**: liquid  
**Category**: text  
**Context**: 128K tokens  
**Pricing**: $0.05 / $0.70 per 1M · billed by your provider (BYOK)  
**Released**: 2026-10-13  
**Web page**: https://anyrouter.dev/model/liquid/lfm-2.5-24b-a2b

**Input modalities**: text  
**Output modalities**: text  

**Tokenizer**: Other

**Capabilities**: chat, streaming, reasoning, function-calling, coding

**Supported parameters**: max_tokens, temperature, top_p, stop, tools, tool_choice, reasoning, response_format

**Aliases**: `liquid/lfm2.5-24b-a2b`

## Usage

**Endpoint**: `POST https://anyrouter.dev/api/v1/chat/completions`

```bash
curl https://anyrouter.dev/api/v1/chat/completions \
  -H "Authorization: Bearer $ANYROUTER_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
  "model": "liquid/lfm-2.5-24b-a2b",
  "messages": [
    {
      "role": "user",
      "content": "Say hi in 3 words."
    }
  ]
}'
```

## Providers

| Provider | Input | Output |
| --- | --- | --- |
| openrouter-byok | $0.05/M (list) | $0.70/M (list) |
| vercel-byok | $0.05/M (list) | $0.70/M (list) |

BYOK routes are billed by your own provider key — AnyRouter charges $0. Rates marked (list) are the provider's published list price.

## Model card

- Architecture: mixture of experts [1]
- Total parameters: 24B [1]
- Active parameters: 2.45B [1]
- Native context: 131072 tokens [1]
- License: LFM Open License v1.0, open weights (https://huggingface.co/LiquidAI/LFM2.5-24B-A2B) [1]

### Strengths [1,2]

- Competitive with a larger MoE on agentic and tool-use benchmarks
- Reasoning at 2B active parameters keeps each thinking token cheap

### Use cases [1]

- Fast agents and tool calling
- Low-cost reasoning for high-volume workloads

### Links

- announcement: https://www.liquid.ai/blog/lfm2-5-24b-a2b
- model card: https://huggingface.co/LiquidAI/LFM2.5-24B-A2B

### Sources

1. [LFM2.5-24B-A2B model card](https://huggingface.co/LiquidAI/LFM2.5-24B-A2B) (accessed 2026-10-17)
2. [LFM2.5-24B-A2B: Built for Fast On-Device Agents](https://www.liquid.ai/blog/lfm2-5-24b-a2b) (accessed 2026-10-17)

## Source

- Upstream docs: https://www.liquid.ai/blog/lfm2-5-24b-a2b