Measured model
Measured on a Mac Studio (M5 Max, 36GB)LFM2-24B-A2B
Liquid AI · Ollama name hf.co/LiquidAI/LFM2-24B-A2B-GGUF:Q4_K_M
On a Mac Studio (M5 Max, 36GB), LFM2-24B-A2B answers a simple question in about 0.1 s, gets 53% of the Japanese knowledge questions right, 66% of the math problems, and uses 14.8GB of memory.
Results
| Measure | Result | Among measured models | Note |
|---|---|---|---|
| Reply time | about 0.1 s | 1 of 33 | Until one simple common-sense question is answered (median of 50) |
| Knowledge (high school to university) | 53% | 31 of 33 | 200 questions, ±7, 2 cut off (counted wrong) |
| Math (word problems) | 66% | 30 of 33 | 100 questions, ±10, 1 cut off (counted wrong) |
| Memory used | 14.8GB | 20 of 33 | |
| EN→JA translation | 32 pts | 11 of 30 | out of 100 (chrF), about 0.3 s a sentence |
| JA→EN translation | 56 pts | 7 of 30 | out of 100 (chrF), about 0.3 s a sentence |
"Among measured models" is the place among the 33 models measured on this site (lower is better for time and memory).
Try it on your Mac
- Install the Mac version of Ollama from ollama.com and start it.
- Type this in Terminal. The first time, the model downloads; then you can chat with it.
ollama run hf.co/LiquidAI/LFM2-24B-A2B-GGUF:Q4_K_M
It used 14.8GB here. To use it comfortably alongside other apps, a Mac with 32GB of memory or more is a good guide (the model then takes about half the memory).
About the model
- Maker
- Liquid AI
- Size
- 23.8B parameters
- Quantization
- Q4_K_M
- Thinks before answering
- Yes
- Longest input at once
- 128,000 tokens
- Released (Hugging Face)
- Feb 24, 2026
- Generation speed
- 160.3 tok/s
- Start-up (loading)
- about 1 s
- Measured on
- Oct 10, 2026
- Ollama
- 0.34.3
- Hugging Face
- LiquidAI/LFM2-24B-A2B
- Series
- LFM
What the parts of a name mean (size, quantization and so on): how to read a model name.
Models of similar memory
- LFM2.5-24B-A2B (14.9GB of memory, knowledge 81%, reply about 7.1 s)
- gpt-oss:20b (12.8GB of memory, knowledge 83%, reply about 1.8 s)
- gemma3:27b (17.5GB of memory, knowledge 76%, reply about 0.4 s)
- muse-glimmer:30b (17.6GB of memory, knowledge 83%, reply about 15 s)
- qwen3.8:27b (18.2GB of memory, knowledge 87%, reply about 7.7 s)
Compare the main models on the text page. The series page lists every model measured.
Task recommendations and capabilities · Measurements by model