Which GPUs and Macs run LFM3-1.7B?
Key facts
Measured downloads are 1.10 GB Q4_K_M, 1.05 GB QAD-Q4_0, 1.43 GB Q6_K, 1.85 GB Q8_0, 3.47 GB BF16 and 0.97 GB MLX-4bit. Yes. LFM3-1.7B needs about 2.0 GB of graphics memory, so it runs on every graphics card setup on this page, even the RTX 3060 12GB.
About 2.0 GB at Q4_K_M, our estimate. Checked . LiquidAI Q4_K_M GGUF.
The cheapest graphics card on our list that runs it (it still needs the rest of a PC): GeForce RTX 3060 12GB, $292€261, used (median of sold listings).
Fast~222 tok/sAnswers appear faster than you can read them.Our price is for a used card. These shops mostly sell new ones, so their price can be higher.
EU shops · Shown to visitors in Europe.
US shops · Shown to visitors outside Europe.
If you buy through these links we may earn a commission. It doesn't change our ranking. How shop links work
As an Amazon Associate I earn from qualifying purchases.
Als Amazon-Partner verdiene ich an qualifizierten Verkäufen.
Als Amazon-Partner verdien ik aan in aanmerking komende aankopen.
We never pass off an estimate as a benchmark. Commissions never change which hardware we recommend. How we calculate
The cheapest complete machine that runs it: Mac mini M4 16GB, about $615€549 (used). Fast~47 tok/s
LIVE IN BROWSER · $0 · NO LOGIN
Test comparable open models (Recursive Flint 40B vs Gemma 4 31B) free in your browser
Open full 8-model playground ↗Click a starter prompt below or type your own question to watch Recursive Flint stream side by side against Gemma 4 31B right on this page:
Get a 1-line alert when a faster quant for LFM3-1.7B drops — or when GPUs that run it drop >10%
Get a 1-line alert when a new open-weight model fits 8 GB VRAM (2.0 GB at Q4 for LFM3-1.7B) or when used RTX 3090 / 4090 & Mac Studio prices drop >10% — weekly, zero spam.
Can I run LFM3-1.7B on a laptop, MacBook Air, phone, or 4 GB–8 GB GPU?
Yes. LFM3-1.7B is 1.73B parameters (our estimate, Lfm3ForCausalLM). The measured Q4_K_M GGUF is 1.10 GB (1.05 GB for QAD-Q4_0) and Apple Silicon MLX-4bit is 0.97 GB. At 8,192 tokens our estimate is 2.01 GB total for Q4_K_M: the file, plus 0.11 GB of conversation memory and 0.80 GB of working space. MLX-4bit totals 1.88 GB on the same estimate.
- Phones and laptops with no discrete GPU: Liquid AI states that at 4-bit precision the model takes about 1 GB and fits a phone or laptop. Our running total is 2.01 GB at Q4_K_M and 2.76 GB at Q8_0, so both fit in 8 GB of system RAM. A 32 GB iGPU laptop, such as the ThinkPad in Fabio Matricardi's CPU lab, has room for either total. Danylo Chernov's Hetzner CPU bench is a write-up of small models on ordinary cloud CPUs.
- Apple Silicon: MLX-4bit (1.88 GB total) and MLX-bf16 (4.37 GB total) fit an 8 GB Mac, which leaves 5.3 GB usable under the same two-thirds graphics share this site uses for Macs up to 32 GB. Mac mini M4 16GB is about €549 ($615), estimated at 47 tok/s at Q4.
- 4 GB, 6 GB and 8 GB GPUs: this list has no 4 GB or 8 GB card. Under the same 0.6 GB reserve, a 4 GB GPU leaves 3.4 GB usable and holds Q4_K_M through Q8_0. BF16 at 4.38 GB needs a 6 GB GPU (5.4 GB usable) or an 8 GB GPU (7.4 GB usable). GeForce RTX 3060 12GB is €261 ($292), estimated at 222 tok/s at Q4.
- Compare 10 open models in the browser (Gemma 4 31B, Recursive Flint, Nemotron 3.5 Lightning), with no login and no local graphics memory. Open the browser test.
Swipe the table sideways to see every column.
| Quality | Measured file | Total at 8,192 tokens | 4 GB GPU (3.4 GB usable) | 8 GB GPU (7.4 GB usable) | 8 GB Mac (5.3 GB usable) | 8 GB system RAM | Cheapest machine on our list |
|---|---|---|---|---|---|---|---|
| MLX-4bitApple Silicon | 0.97 GB | 1.88 GB | Apple only | Apple only | Yes | Apple only | Mac mini M4 16GB, $615€549 (used), price unverified |
| QAD-Q4_0GGUF | 1.05 GB | 1.96 GB | Yes | Yes | Yes | Yes | GeForce RTX 3060 12GB, $292€261, used (median of sold listings) |
| Q4_K_MGGUF | 1.10 GB | 2.01 GB | Yes | Yes | Yes | Yes | GeForce RTX 3060 12GB, $292€261, used (median of sold listings) |
| MLX-5bitApple Silicon | 1.19 GB | 2.10 GB | Apple only | Apple only | Yes | Apple only | Mac mini M4 16GB, $615€549 (used), price unverified |
| Q5_K_MGGUF | 1.26 GB | 2.17 GB | Yes | Yes | Yes | Yes | GeForce RTX 3060 12GB, $292€261, used (median of sold listings) |
| MLX-6bitApple Silicon | 1.41 GB | 2.32 GB | Apple only | Apple only | Yes | Apple only | Mac mini M4 16GB, $615€549 (used), price unverified |
| Q6_KGGUF | 1.43 GB | 2.34 GB | Yes | Yes | Yes | Yes | GeForce RTX 3060 12GB, $292€261, used (median of sold listings) |
| MLX-8bitApple Silicon | 1.84 GB | 2.75 GB | Apple only | Apple only | Yes | Apple only | Mac mini M4 16GB, $615€549 (used), price unverified |
| Q8_0GGUF | 1.85 GB | 2.76 GB | Yes | Yes | Yes | Yes | GeForce RTX 3060 12GB, $292€261, used (median of sold listings) |
| MLX-bf16Apple Silicon | 3.46 GB | 4.37 GB | Apple only | Apple only | Yes | Apple only | Mac mini M4 16GB, $615€549 (used), price unverified |
| BF16GGUF | 3.47 GB | 4.38 GB | No | Yes | Yes | Yes | GeForce RTX 3060 12GB, $292€261, used (median of sold listings) |
F16 matches BF16's measured size, so the table lists BF16 once. MLX files run on Apple Silicon only. Conversation memory is 0.11 GB and working space is 0.80 GB on every row. Both are our estimate.
LFM3-1.7B, Spark-X2.6-4B and Nemotron 3.5 Bolt E4B on 8 GB
Totals add the measured file, conversation memory at 8,192 tokens, and 0.80 GB of working space. File sizes are measured. The totals and the speeds are our estimate.
| Model | Parameters | Architecture | Conversation memory | Q4_K_M file to total | BF16 file to total | Max context | Minimum hardware for Q4_K_M |
|---|---|---|---|---|---|---|---|
| LFM3-1.7B | 1.73B dense, our estimate. | 24 layers: 13 gated short convolutions, 5 sliding-window (256), 6 global attention. | 0.11 GB | 1.10 GB measured GGUF → 2.01 GB (0.97 GB MLX → 1.88 GB) | 3.47 GB → 4.38 GB. Fits an 8 GB GPU (7.4 GB usable). | 262,144 | Phone or laptop CPU, or a 4 GB GPU (3.4 GB usable). About 222 tok/s on GeForce RTX 3060 12GB, our estimate. Mac mini M4 16GB, about 47 tok/s, our estimate. |
| Spark-X2.6-4B | 4.11B dense, our estimate. | 36 layers: 27 sliding-window (512) and 9 full attention. | 0.36 GB | 2.62 GB measured GGUF → 3.78 GB. Official Q4_K_M 3.54 GB → 4.70 GB | 8.23 GB → 9.39 GB. Fits a 12 GB GPU (11.4 GB usable). | 1,048,576 | 6 GB GPU (5.4 GB usable) or an 8 GB Mac (5.3 GB usable). About 93 tok/s on GeForce RTX 3060 12GB, our estimate. |
| NVIDIA Nemotron 3.5 Bolt E4B | 8B parameters, our estimate. The card calls 4B of that effective, from per-layer embeddings. | 42 layers: 35 sliding-window (512) and 7 full attention. | 0.27 GB | 6.18 GB measured GGUF → 7.25 GB (4.20 GB MLX → 5.27 GB) | 15.05 GB → 16.12 GB. Fits a 24 GB GPU (23.4 GB usable). | 65,536. The GGUF header and Ollama list 128K. | 8 GB GPU (7.4 GB usable) or a 16 GB Mac (10.7 GB usable). About 43 tok/s on GeForce RTX 3060 12GB, our estimate. |
Which graphics cards run it
Seven popular graphics cards and two PCs with two cards each, at the standard quality (Q4). Click one for the full answer: memory, speed and price.
Runs well10
The whole model fits in the graphics memory.
Which Macs and mini PCs run it
Macs and mini PCs from the picker, at the standard quality (Q4). A Mac shares its memory with the rest of the computer. Each name opens that machine's page.
Runs well12
The whole model fits in the memory the graphics can use.
- Mac mini M4 16GB
- Mac mini M6 32GB
- Mac Studio M5 Max 36GB
- Mac mini M4 Pro 64GB
- Mac mini M5 Pro 64GB
- GMKtec EVO-X2 (Ryzen AI Max+ 395) 64GB
- Mac Studio M5 Ultra 96GB
- GMKtec EVO-X2 (Ryzen AI Max+ 395) 128GB
- GEEKOM A9 Mega (Ryzen AI Max+ 395) 128GB
- Beelink GTR9 Pro (Ryzen AI Max+ 395) 128GB
- Mac Studio M5 Max 128GB
- Mac Studio M5 Ultra 256GB
Cheapest hardware for each size
From all 43 graphics cards, PCs and Macs on our list. "Runs slowly" means part of the model runs from normal memory: cheaper, but much slower. A smaller download needs less memory and answers a little less well.
- Standard quality (Q4_K_M)GeForce RTX 3060 12GB, $292€261, used (median of sold listings)
- Standard quality (Q4_K_M), runs slowlyNo machine on our list runs this size.
- Smallest download (MLX-4bit)Mac mini M4 16GB, about $615€549 (used)
- Smallest download (MLX-4bit), runs slowlyNo machine on our list runs this size.
Prices excl. shipping unless marked. Euro prices incl. VAT where the seller charges it. Dollar prices are converted from our euro prices without the 21% EU VAT, before US sales tax.
Prices moved: see Building a local LLM rig in the RAMageddon.
About LFM3-1.7B
1.73B parameters. Parameter count is an estimate. More parameters usually means a smarter model that needs more memory.
- Licence:
- LFM Open License v1.0 source
- Longest conversation:
- about 393 pages of text (262,144 tokens)
- Input:
- text
- Released:
- source
Sources and method
Dense text model for local agents, 1.73B parameters. Hugging Face safetensors metadata lists 1,728,801,536 parameters, all BF16. 1.73B is that count to two decimals, so our parameter count is an estimate. The model card also states 1.73B. Architecture Lfm3ForCausalLM (model type lfm3): 24 layers, 13 gated short convolutions, 5 sliding-window attention layers with a 256-token window and 6 global attention layers, 4 KV heads, head dim 128, max context 262,144 (the card calls this 256K). Our estimate has only two attention buckets, so the 18 layers that are not global attention, including the 13 convolutions, are counted with that 256-token window. Conversation memory is therefore a little higher than the 6 global and 5 sliding layers alone. At 8,192 tokens that estimate is 0.11 GB of conversation memory plus 0.80 GB of working space. LiquidAI's Q4_K_M measured 1.1 GB on 28 Oct 2026 (1,099,627,808 bytes), so the total at 8,192 tokens is 2.01 GB. Q8_0 measured 1.85 GB (1,846,238,496 bytes) and BF16 measured 3.47 GB (3,466,626,336 bytes). F16 is the same 3,466,626,336 bytes. QAD Q4_0 measured 1.05 GB (1,049,615,680 bytes). Official MLX 4-bit weights measured 0.97 GB (973,083,386 bytes). Official GGUF, MLX and ONNX builds are published. ONNX is not in the fit table. LFM Open License v1.0 allows commercial use below $10 million annual revenue. The card says that at 4-bit precision the model takes about 1 GB and fits a phone or laptop. That is Liquid AI's statement. Our total adds conversation memory and working space to the measured file. The GGUF card says the files are for local inference on CPUs and GPUs, and that llama.cpp needs an LFM3-compatible build.
Ran LFM3-1.7B on your own hardware? Send us your llama-bench result.
Memory at each quality level
Each download of LFM3-1.7B is a different size. The total adds the file, the memory a conversation of about 12 pages of text (8,192 tokens) takes, and 0.8 GB of working space. A measured file size is labelled; the rest is our estimate. The last column is the cheapest machine that holds the total.
Swipe the table sideways to see every column, up to the cheapest machine.
| Quality (file) | File size | Conversation | Working space | Total | Cheapest machine that holds it |
|---|---|---|---|---|---|
| BF16 | 3.47 GBmeasured GGUF | 0.1 GB | 0.8 GB | 4.4 GB | GeForce RTX 3060 12GB, $292€261, used (median of sold listings) |
| F16 | 3.47 GBmeasured GGUF | 0.1 GB | 0.8 GB | 4.4 GB | GeForce RTX 3060 12GB, $292€261, used (median of sold listings) |
| Q8_0 | 1.85 GBmeasured GGUF | 0.1 GB | 0.8 GB | 2.8 GB | GeForce RTX 3060 12GB, $292€261, used (median of sold listings) |
| Q6_K | 1.43 GBmeasured GGUF | 0.1 GB | 0.8 GB | 2.3 GB | GeForce RTX 3060 12GB, $292€261, used (median of sold listings) |
| Q5_K_M | 1.26 GBmeasured GGUF | 0.1 GB | 0.8 GB | 2.2 GB | GeForce RTX 3060 12GB, $292€261, used (median of sold listings) |
| Q4_K_M | 1.10 GBmeasured GGUF | 0.1 GB | 0.8 GB | 2.0 GB | GeForce RTX 3060 12GB, $292€261, used (median of sold listings) |
| QAD-Q4_0 | 1.05 GBmeasured GGUF | 0.1 GB | 0.8 GB | 2.0 GB | GeForce RTX 3060 12GB, $292€261, used (median of sold listings) |
| MLX-bf16MLX: Apple Silicon only | 3.46 GBmeasured MLX | 0.1 GB | 0.8 GB | 4.4 GB | Mac mini M4 16GB, $615€549 (used), price unverified |
| MLX-8bitMLX: Apple Silicon only | 1.84 GBmeasured MLX | 0.1 GB | 0.8 GB | 2.8 GB | Mac mini M4 16GB, $615€549 (used), price unverified |
| MLX-6bitMLX: Apple Silicon only | 1.41 GBmeasured MLX | 0.1 GB | 0.8 GB | 2.3 GB | Mac mini M4 16GB, $615€549 (used), price unverified |
| MLX-5bitMLX: Apple Silicon only | 1.19 GBmeasured MLX | 0.1 GB | 0.8 GB | 2.1 GB | Mac mini M4 16GB, $615€549 (used), price unverified |
| MLX-4bitMLX: Apple Silicon only | 0.97 GBmeasured MLX | 0.1 GB | 0.8 GB | 1.9 GB | Mac mini M4 16GB, $615€549 (used), price unverified |
Speeds and memory are our estimate from the hardware specs, not a measured test. Prices are dated and may change. How we calculate
Questions
- How much memory does LFM3-1.7B need at Q4_K_M, Q8_0 and BF16, and does it fit a 4 GB GPU, laptop or phone?
- At Q4_K_M (1.1 GB measured GGUF + 0.11 GB conversation memory at 8,192 tokens + 0.80 GB working space = 2.01 GB / 2.0 GB total estimate). At Q8_0 (1.85 GB measured GGUF + 0.11 GB conversation memory at 8,192 tokens + 0.80 GB working space = 2.76 GB / 2.8 GB total estimate). At BF16 (3.47 GB measured GGUF + 0.11 GB conversation memory at 8,192 tokens + 0.80 GB working space = 4.38 GB / 4.4 GB total estimate). The QAD Q4_0 file is QAD-Q4_0 (1.05 GB measured GGUF, 2.0 GB total estimate). A 4 GB card would leave 3.4 GB usable under the same 0.6 GB reserve this site uses on a graphics card (this list has no 4 GB card). Q4_K_M fits that 3.4 GB. BF16 at 4.4 GB does not fit that 3.4 GB. GeForce RTX 3060 12GB (11.4 GB usable) holds the Q4_K_M file, and so does every larger card on https://localllmrig.com/can-i-run/lfm3-1.7b. The model card says that at 4-bit precision it takes about 1 GB of memory and fits a phone or laptop. That is Liquid AI's statement. Our total adds conversation memory and working space to the measured file. MLX-4bit (0.97 GB measured MLX, 1.9 GB total estimate) is the official Apple Silicon build.
- Can I run LFM3-1.7B on a MacBook Air or Mac mini M4, and should I use MLX or GGUF?
- Yes. Liquid AI publishes MLX weights for Apple Silicon and GGUF files for llama.cpp, Ollama and LM Studio. MLX-4bit is 0.97 GB measured (1.88 GB total estimate at 8,192 tokens). MLX-bf16 is 3.46 GB measured (4.37 GB total estimate). An 8 GB Mac leaves 5.3 GB usable under the same two-thirds graphics share this site uses for Macs up to 32 GB, so both of those totals fit. Mac mini M4 16GB leaves 10.7 GB usable of 16 GB (about €549 ($615), about 47 tok/s estimate at Q4). Use MLX on Apple Silicon. Use GGUF when the runtime is llama.cpp, Ollama or LM Studio. The GGUF card says llama.cpp needs an LFM3-compatible build.
- How does LFM3-1.7B compare with Spark-X2.6-4B and NVIDIA Nemotron 3.5 Bolt E4B on 8 GB hardware?
- All three fit an 8 GB GPU (7.4 GB usable) at their standard 4-bit file. LFM3-1.7B (1.73B parameters, our estimate) needs 2.01 GB total at Q4_K_M (1.10 GB measured GGUF) and 4.38 GB at BF16, so BF16 also fits that 7.4 GB. Spark-X2.6-4B (4.11B parameters, our estimate) needs 3.78 GB total at bartowski Q4_K_M (2.62 GB measured GGUF) and 4.70 GB for the official 3.54 GB Q4_K_M file. Its Q8_0 total is 5.54 GB. Its BF16 file is 8.23 GB (9.39 GB total), which fits a 12 GB GPU (11.4 GB usable). On GeForce RTX 3060 12GB the Q4 speed estimate is about 93 tok/s. NVIDIA Nemotron 3.5 Bolt E4B is 8B parameters. The card calls 4B of that effective, from per-layer embeddings. Its measured Q4_K_M file is 6.18 GB (7.25 GB total at 8,192 tokens), which fits the 7.4 GB an 8 GB GPU leaves usable. Its BF16 file is 15.05 GB (16.12 GB total), which fits a 24 GB GPU (23.4 GB usable). The context stored for it is 65,536 tokens. The GGUF header and Ollama list 128K.
- Why is LFM3-1.7B's conversation memory 0.11 GB at 8,192 tokens when the max context is 262,144?
- LFM3-1.7B uses Lfm3ForCausalLM (lfm3): 24 layers, 13 gated short convolutions, 5 sliding-window attention layers with a 256-token window, and 6 global attention layers (4 KV heads, head dim 128). Our estimate has two buckets, so the 18 layers that are not global attention, including the 13 convolutions, are counted with that 256-token window. Conversation memory is therefore a little higher than the 6 global and 5 sliding layers alone. At 8,192 tokens that estimate is 0.11 GB. At the full 262,144-token context it is 3.23 GB.
- Does Lfm3ForCausalLM run in llama.cpp, Ollama, LM Studio, MLX, ONNX, vLLM or Transformers?
- The LFM3-1.7B model card lists GGUF for llama.cpp, Ollama and LM Studio, MLX for Apple Silicon, ONNX for ONNX Runtime, and BF16 weights for vLLM, SGLang and Transformers. The GGUF card says llama.cpp needs an LFM3-compatible build, and that the GGUF files are for local inference on CPUs and GPUs. The Transformers example on the model card sets trust_remote_code because the repository ships custom code. ONNX is published and is not in the fit table on https://localllmrig.com/can-i-run/lfm3-1.7b.
- Where can I test open models in the browser before downloading LFM3-1.7B?
- LFM3-1.7B is not one of the 10 free models. You can test those open models in the browser at https://localllmrig.com/ with no login and no local graphics memory, then read the fit on https://localllmrig.com/can-i-run/lfm3-1.7b.
- What licence does LFM3-1.7B use, and is commercial use allowed?
- LFM3-1.7B is listed as LFM Open License v1.0. Our reading of the licence is that commercial use is allowed below $10 million annual revenue, and is not licensed at or above that. That is our reading of the licence, not legal advice. File sizes and the machines that hold them are on https://localllmrig.com/can-i-run/lfm3-1.7b. You can try open models live for free at https://localllmrig.com/.