LFM3 1.7B — Hardware Requirements & GPU Compatibility
ChatFunctions
LFM3-1.7B is Liquid AI's text-only language model for local agents and specialized applications, and the smallest instruct model in the LFM3 family. It has 1.73 billion parameters and a tri-hybrid architecture of 24 layers: 13 gated short convolutions, 5 sliding-window attention layers, and 6 global attention layers. It is post-trained for agentic harnesses such as Hermes Agent, Pi, and OpenClaw, with tool calling and structured outputs, and is meant to be fine-tuned to a specific domain on a single GPU. The publisher advises using retrieval or search for knowledge-heavy tasks. Context length is 256K tokens, input plus output. It is released under the LFM Open License v1.0, a custom license whose terms are in the repository, and was published in October 2026. At 4-bit precision it needs about 1 GB of memory, so it runs on almost any modern phone, laptop, or consumer GPU.
Runs great on a 24 GB GPU
Runs from 8 GB (Q4_K_M · 128K)Best on 24 GB (Q4_K_M · ~455 tok/s · 256K)Macs from 8 GB (Q4_K_M · ~50 tok/s)
Other
Reference GPU · RTX 4090 24 GB
Estimates, ±20% · how we estimate
Run LFM3 1.7B
One recommended way. Every other method, OS and quant is in the step-by-step guide.Each way opens a step-by-step guide for your OS and quant.
Recommended
ollama run hf.co/LiquidAI/LFM3-1.7B-GGUF:Q4_K_MQ4_K_M fits fully on the RTX 4090 · 21.9 GB free · 16K context
Download: LFM3-1.7B-Q4_K_M.gguf1.10 GBLiquidAI
Ways to runOther ways
each opens step-by-stepRun in cloud
Runs on RTX 3060 12GB (9 GB headroom) · Q4_K_M
- Generation speed
- ~163 tok/s
- generation speed
- Cost per 1M output tokens
- $0.11
- per 1M output tokens
How Much VRAM Does LFM3 1.7B Need?
Select a quantization level; the listed hardware adheres to it.
| Quantization | VRAM | +32K | On RTX 4090 (reference) | |
|---|---|---|---|---|
| 1.1 GB | 2.7 GB | 2-bit · K-quant — 2-bit quantization with K-quant improvements | Fits: 256K context · ~575 tok/s | |
| 1.2 GB | 2.7 GB | 3-bit small — 3-bit small quantization | Fits: 256K context · ~565 tok/s | |
| 1.2 GB | 2.8 GB | 3-bit medium — 3-bit medium quantization | Fits: 256K context · ~528 tok/s | |
| 1.3 GB | 2.8 GB | 4-bit · legacy — 4-bit legacy quantization | Fits: 256K context · ~516 tok/s | |
| 1.4 GB | 3 GB | 4-bit medium — 4-bit medium quantization — most popular sweet spot | Fits: 256K context · ~455 tok/s | |
| 1.6 GB | 3.1 GB | 5-bit medium — 5-bit medium quantization — good quality/size tradeoff | Fits: 256K context · ~402 tok/s | |
| 1.8 GB | 3.3 GB | 6-bit, very good — 6-bit quantization, very good quality | Fits: 256K context · ~358 tok/s | |
| 2.1 GB | 3.6 GB | 8-bit, near-lossless — 8-bit quantization, near-lossless | Fits: 256K context · ~308 tok/s |
Selected Q4_K_M
4-bit medium quantization2.1GBat 16K context
- Weights
- 1.0 GB
- KV cache
- 0.8 GB
- Overhead
- 0.3 GB
Fits RTX 4090 (reference) · 21.9 GB free · ~455 tok/s
≥50 tok/s needs ≥111 GB/s · RTX 5090 ~809, RTX 3090 Ti ~455
Which Hardware Can Run LFM3 1.7B?
Q4_K_M · 1.4 GBLFM3 1.7B (Q4_K_M) requires 1.4 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 262K context window can add up to 12.8 GB, bringing total usage to 14.2 GB. 52 GPUs can run it, including NVIDIA GeForce RTX 5090, NVIDIA GeForce RTX 3090 Ti. 163 devices with unified memory can run LFM3 1.7B, including Mac Studio (M3 Ultra, 256GB), Mac Studio (M3 Ultra, 512GB).
≤ 11 GB
≤ 11 GB tier →fits · 5.9 GB free · up to 128K ctx
16 GPUs · 1 Mac class · 2 device families
RTX 3080~343Simulate output on RTX 3080RTX 2080 Ti~278Simulate output on RTX 2080 TiRTX 3070 Ti~275Simulate output on RTX 3070 TiM3*8 GB~50Simulate output on iMac 24" M3 (8 GB)Orin Nano~46Simulate output on NVIDIA Jetson Orin Nano 8GB (Super)iPhone~33Simulate output on iPhone 1720 GB
fits · 17.9 GB free · up to 256K ctx
1 GPU
RX 7900 XT~333Simulate output on RX 7900 XT32–40 GB
fits · 29.9 GB free · up to 256K ctx
1 GPU · 3 Mac classes · 4 device families
RTX 5090~809Simulate output on RTX 5090M5 Max36 GB~224Simulate output on Mac Studio M5 Max (36 GB)M532 GB~75Simulate output on MacBook Air 13" M5 (32 GB)Orin~92Simulate output on NVIDIA Jetson AGX Orin 32GBSnapdragon X Elite (X1E-80-100)~61Simulate output on Snapdragon X Elite Copilot+ PC
By VRAM tier · typical GPUs and Macs at Q4_K_M
Alternatives to LFM3 1.7B
LFM3 family · 3 models →Similar models
6| Model | Score | VRAM | tok/s | Why it is similar |
|---|---|---|---|---|
| Granite 4.0 1BIBM | — | 2.6 GB | ~452 | same size |
| Qwen 1 8BAlibaba | — | 1.9 GB | ~542 | same size |
| JiRackUltra 1BCMSManhattan | — | 1.8 GB | ~461 | same size |
| Qwen3 1.7B BaseAlibaba | — | 3.2 GB | ~417 | same size |
| DeepSeek R1 Distill Qwen 1.5BDeepSeek | — | 1.8 GB | ~461 | same size |
| SmolLM2 1.7B InstructHugging Face | — | 2.9 GB | ~379 | same size |
Specifications
- Publisher
- Liquid AI
- Family
- LFM3
- Parameters
- 1.7B
- Architecture
- Lfm3ForCausalLM
- Context Length
- 262,144 tokens
- Vocabulary Size
- 128,000
- Release Date
- 2026-10-23
- License
- Other
Frequently Asked Questions
How much VRAM does LFM3 1.7B need?
LFM3 1.7B requires 1.4 GB of VRAM at Q4_K_M, or 3.9 GB at BF16. Full 262K context adds up to 12.8 GB (14.2 GB total).
VRAM = Weights + KV Cache + Overhead
Weights = 1.7B × 4.8 bits ÷ 8 = 1 GB
KV Cache + Overhead ≈ 0.4 GB (at 2K context + ~0.3 GB framework)
Fit ratings and hardware model lists check this model with room for a 16K-token context, which needs a little more memory.
KV Cache + Overhead ≈ 13.2 GB (at full 262K context)
VRAM usage by quantization
What's the best quantization for LFM3 1.7B?
For LFM3 1.7B, Q4_K_M (1.4 GB) offers the best balance of quality and VRAM usage. Q4_K_L (1.5 GB) provides better quality if you have the VRAM. The smallest option is IQ2_XS at 0.9 GB.
VRAM requirement by quantization
★ Recommended — best balance of quality and VRAM usage.
Can I run LFM3 1.7B on a Mac?
Yes — MacBook Air 13" M3 (8 GB) and 145 other Macs can run LFM3 1.7B. Apple Silicon uses unified memory, so the model shares RAM with the system. At Q4_K_M you need at least 1.4 GB of usable unified memory (RAM minus macOS overhead).
Can I run LFM3 1.7B locally?
Yes — LFM3 1.7B can run locally on consumer hardware. Popular tools include Ollama, LM Studio, and llama.cpp.
How fast is LFM3 1.7B?
At Q4_K_M, LFM3 1.7B can reach ~3333 tok/s on AMD Instinct MI350X. On NVIDIA GeForce RTX 4090: ~455 tok/s. Speed depends mainly on GPU memory bandwidth. Real-world results typically within ±20%.
tok/s = (bandwidth GB/s ÷ model GB) × efficiency
Example: NVIDIA B200 → 8000 ÷ 1.44 × 0.65 = ~3611 tok/s
Estimated speed at Q4_K_M (1.4 GB)
Real-world results typically within ±20%. Speed depends on batch size, quantization kernel, and software stack.
What's the download size of LFM3 1.7B?
At Q4_K_M, the download is about 1.04 GB. The full-precision BF16 version is 3.46 GB. The smallest option (IQ2_XS) is 0.52 GB.
Which GPUs can run LFM3 1.7B?
52 consumer GPUs can run LFM3 1.7B at Q4_K_M. Top options include AMD Radeon RX 6700 XT, AMD Radeon RX 6800, AMD Radeon RX 6800 XT. 52 GPUs have plenty of headroom for comfortable inference.
Which devices can run LFM3 1.7B?
163 devices with unified memory can run LFM3 1.7B at Q4_K_M (1.4 GB), including AMD Ryzen AI 9 HX 370 (Strix Point) Laptop, ASUS Ascent GX10, Beelink GTR9 Pro (Ryzen AI Max+ 395, 128 GB), Framework Desktop (Ryzen AI Max+ 395, 128 GB). Apple Silicon Macs use unified memory shared between CPU and GPU, making them well-suited for local LLM inference.