LFM2.5 1.2B Instruct — Hardware Requirements & GPU Compatibility
Chat
LFM2.5 1.2B Instruct is an instruction-tuned model from Liquid AI that uses a novel hybrid architecture combining state-space models with attention mechanisms. At just 1.2 billion parameters, it is exceptionally lightweight and can run on virtually any hardware, including laptops and edge devices. Liquid AI's unconventional architecture aims to deliver better efficiency and longer context handling than traditional transformer models at this scale, making it an interesting option for users exploring alternatives to standard transformer-based LLMs.
219.2K downloads
527 likesSpecifications
- Publisher
- LiquidAI
- Parameters
- 1.2B
- Architecture
- Lfm2ForCausalLM
- Context Length
- 128,000 tokens
- Vocabulary Size
- 65,536
- Release Date
- 2026-02-24
- License
- Other
Get Started
HuggingFace
How Much VRAM Does LFM2.5 1.2B Instruct Need?
Select a quantization to see compatible GPUs below.
| Quantization | Bits | VRAM | + Context | File Size | Quality |
|---|---|---|---|---|---|
| Q4_0 | 4.00 | 0.9 GB | 5.1 GB | 0.59 GB | 4-bit legacy quantization |
| Q4_K_M | 4.80 | 1.1 GB | 5.2 GB | 0.70 GB | 4-bit medium quantization — most popular sweet spot |
| Q5_K_M | 5.70 | 1.2 GB | 5.3 GB | 0.83 GB | 5-bit medium quantization — good quality/size tradeoff |
| Q6_K | 6.60 | 1.3 GB | 5.5 GB | 0.97 GB | 6-bit quantization, very good quality |
| Q8_0 | 8.00 | 1.5 GB | 5.7 GB | 1.17 GB | 8-bit quantization, near-lossless |
Which GPUs Can Run LFM2.5 1.2B Instruct?
Q4_K_M · 1.1 GBLFM2.5 1.2B Instruct (Q4_K_M) requires 1.1 GB of VRAM to load the model weights. For comfortable inference with headroom for KV cache and system overhead, 2+ GB is recommended. Using the full 128K context window can add up to 4.1 GB, bringing total usage to 5.2 GB. 35 GPUs can run it, including NVIDIA GeForce RTX 5090, NVIDIA GeForce RTX 3090 Ti.
Runs great
— Plenty of headroomWhich Devices Can Run LFM2.5 1.2B Instruct?
Q4_K_M · 1.1 GB33 devices with unified memory can run LFM2.5 1.2B Instruct, including NVIDIA DGX H100, NVIDIA DGX A100 640GB.
Runs great
— Plenty of headroomRelated Models
Frequently Asked Questions
- How much VRAM does LFM2.5 1.2B Instruct need?
LFM2.5 1.2B Instruct requires 1.1 GB of VRAM at Q4_K_M, or 1.5 GB at Q8_0. Full 128K context adds up to 4.1 GB (5.2 GB total).
VRAM = Weights + KV Cache + Overhead
Weights = 1.2B × 4.8 bits ÷ 8 = 0.7 GB
KV Cache + Overhead ≈ 0.4 GB (at 2K context + ~0.3 GB framework)
KV Cache + Overhead ≈ 4.5 GB (at full 128K context)
VRAM usage by quantization
Q4_K_M1.1 GBQ4_K_M + full context5.2 GB- What's the best quantization for LFM2.5 1.2B Instruct?
For LFM2.5 1.2B Instruct, Q4_K_M (1.1 GB) offers the best balance of quality and VRAM usage. Q5_K_M (1.2 GB) provides better quality if you have the VRAM. The smallest option is Q4_0 at 0.9 GB.
VRAM requirement by quantization
Q4_00.9 GB~85%Q4_K_M ★1.1 GB~89%Q5_K_M1.2 GB~92%Q6_K1.3 GB~95%Q8_01.5 GB~99%★ Recommended — best balance of quality and VRAM usage.
- Can I run LFM2.5 1.2B Instruct on a Mac?
LFM2.5 1.2B Instruct requires at least 0.9 GB at Q4_0, which exceeds the unified memory of most consumer Macs. You would need a Mac Studio or Mac Pro with a high-memory configuration.
- Can I run LFM2.5 1.2B Instruct locally?
Yes — LFM2.5 1.2B Instruct can run locally on consumer hardware. At Q4_K_M quantization it needs 1.1 GB of VRAM. Popular tools include Ollama, LM Studio, and llama.cpp.
- How fast is LFM2.5 1.2B Instruct?
At Q4_K_M, LFM2.5 1.2B Instruct can reach ~2724 tok/s on AMD Instinct MI300X. On NVIDIA GeForce RTX 4090: ~612 tok/s. Speed depends mainly on GPU memory bandwidth. Real-world results typically within ±20%.
tok/s = (bandwidth GB/s ÷ model GB) × efficiency
Example: AMD Instinct MI300X → 5300 ÷ 1.1 × 0.55 = ~2724 tok/s
Estimated speed at Q4_K_M (1.1 GB)
AMD Instinct MI300X~2724 tok/sNVIDIA GeForce RTX 4090~612 tok/sNVIDIA H100 SXM~2036 tok/sAMD Instinct MI250X~1684 tok/sReal-world results typically within ±20%. Speed depends on batch size, quantization kernel, and software stack.
- What's the download size of LFM2.5 1.2B Instruct?
At Q4_K_M, the download is about 0.70 GB. The full-precision Q8_0 version is 1.17 GB. The smallest option (Q4_0) is 0.59 GB.