LFM3 1.7B VRAM requirements: which GPUs can run it locally?
1.73B params 256K max context Tool useReasoning
The smallest GPU memory that runs LFM3 1.7B well is 2 GB (Q8_0, e.g. T400). 132 of 132 GPUs run it well.
SWE-bench Verified 43.4
VRAM needed per quantization
| Quantization | Weights | 16K context | 32K context | Full 256K |
|---|---|---|---|---|
| Q4_K_M | 1.0 GB | 1.2 GB | 1.4 GB | 3.8 GB |
| Q8_0 | 1.7 GB | 1.9 GB | 2.1 GB | 4.5 GB |
Weights plus fp16 KV cache. Leave ~1 GB headroom for the inference engine.
Run it locally
Each command downloads the GGUF from Hugging Face, then Ollama starts a chat and llama.cpp starts an OpenAI-compatible server with a web UI on port 8080.
Ollama
ollama run hf.co/LiquidAI/LFM3-1.7B-GGUF:Q4_K_M llama.cpp
llama-server -hf LiquidAI/LFM3-1.7B-GGUF:Q4_K_M Ollama
ollama run hf.co/LiquidAI/LFM3-1.7B-GGUF:Q8_0 llama.cpp
llama-server -hf LiquidAI/LFM3-1.7B-GGUF:Q8_0 Cheapest GPUs that can run LFM3 1.7B
Lowest launch MSRP among GPUs that run its smallest quantization (Q4_K_M) well, with at least 16K context.
As an Amazon Associate, I earn from qualifying purchases.
GPUs that can run LFM3 1.7B
Best quantization that fits on each GPU, with the context and speed it gets. For chips with several memory sizes, the smallest that works is shown.
NVIDIA
- T400 2 GB Q8_0 25K context · ~28 tok/s
- T1000 4 GB Q8_0 207K context · ~56 tok/s
- T400 4GB 4 GB Q8_0 207K context · ~28 tok/s
- T600 4 GB Q8_0 207K context · ~56 tok/s
- GTX 1060 6GB 6 GB Q8_0 256K context · ~67 tok/s
- GTX 1660 6 GB Q8_0 256K context · ~67 tok/s
- GTX 1660 SUPER 6 GB Q8_0 256K context · ~118 tok/s
- GTX 1660 Ti 6 GB Q8_0 256K context · ~101 tok/s
- RTX 2060 6 GB Q8_0 256K context · ~118 tok/s
- RTX 3050 6GB 6 GB Q8_0 256K context · ~59 tok/s
- RTX A2000 6 GB Q8_0 256K context · ~101 tok/s
- GTX 1070 8 GB Q8_0 256K context · ~90 tok/s
- GTX 1070 Ti 8 GB Q8_0 256K context · ~90 tok/s
- GTX 1080 8 GB Q8_0 256K context · ~112 tok/s
- RTX 2060 SUPER 8 GB Q8_0 256K context · ~157 tok/s
- RTX 2070 8 GB Q8_0 256K context · ~157 tok/s
- RTX 2070 SUPER 8 GB Q8_0 256K context · ~157 tok/s
- RTX 2080 8 GB Q8_0 256K context · ~157 tok/s
- RTX 2080 SUPER 8 GB Q8_0 256K context · ~174 tok/s
- RTX 3050 8 GB Q8_0 256K context · ~79 tok/s
- RTX 3060 Ti 8 GB Q8_0 256K context · ~157 tok/s
- RTX 3060 Ti GDDR6X 8 GB Q8_0 256K context · ~214 tok/s
- RTX 3070 8 GB Q8_0 256K context · ~157 tok/s
- RTX 3070 Ti 8 GB Q8_0 256K context · ~214 tok/s
- RTX 4060 8 GB Q8_0 256K context · ~96 tok/s
- RTX 4060 Ti 8 GB Q8_0 256K context · ~101 tok/s
- RTX 5050 8 GB Q8_0 256K context · ~112 tok/s
- RTX 5060 8 GB Q8_0 256K context · ~157 tok/s
- RTX 5060 Ti 8 GB Q8_0 256K context · ~157 tok/s
- T1000 8GB 8 GB Q8_0 256K context · ~56 tok/s
- RTX 3080 10 GB Q8_0 256K context · ~267 tok/s
- GTX 1080 Ti 11 GB Q8_0 256K context · ~170 tok/s
- RTX 2080 Ti 11 GB Q8_0 256K context · ~216 tok/s
- RTX 2060 12GB 12 GB Q8_0 256K context · ~118 tok/s
- RTX 3060 12 GB Q8_0 256K context · ~126 tok/s
- RTX 3080 12GB 12 GB Q8_0 256K context · ~320 tok/s
- RTX 3080 Ti 12 GB Q8_0 256K context · ~320 tok/s
- RTX 4070 12 GB Q8_0 256K context · ~177 tok/s
- RTX 4070 SUPER 12 GB Q8_0 256K context · ~177 tok/s
- RTX 4070 Ti 12 GB Q8_0 256K context · ~177 tok/s
- RTX 5070 12 GB Q8_0 256K context · ~236 tok/s
- RTX A2000 12GB 12 GB Q8_0 256K context · ~101 tok/s
- RTX 4060 Ti 16GB 16 GB Q8_0 256K context · ~101 tok/s
- RTX 4070 Ti SUPER 16 GB Q8_0 256K context · ~236 tok/s
- RTX 4080 16 GB Q8_0 256K context · ~252 tok/s
- RTX 4080 SUPER 16 GB Q8_0 256K context · ~259 tok/s
- RTX 5060 Ti 16GB 16 GB Q8_0 256K context · ~157 tok/s
- RTX 5070 Ti 16 GB Q8_0 256K context · ~315 tok/s
- RTX 5080 16 GB Q8_0 256K context · ~337 tok/s
- RTX 2000 Ada 16 GB Q8_0 256K context · ~79 tok/s
- RTX A4000 16 GB Q8_0 256K context · ~157 tok/s
- RTX PRO 2000 Blackwell 16 GB Q8_0 256K context · ~101 tok/s
- Tesla T4 16 GB Q8_0 256K context · ~112 tok/s
- Tesla V100 16GB 16 GB Q8_0 256K context · ~316 tok/s
- RTX 4000 Ada 20 GB Q8_0 256K context · ~126 tok/s
- RTX 3090 24 GB Q8_0 256K context · ~329 tok/s
- RTX 3090 Ti 24 GB Q8_0 256K context · ~354 tok/s
- RTX 4090 24 GB Q8_0 256K context · ~354 tok/s
- RTX 4500 Ada 24 GB Q8_0 256K context · ~152 tok/s
- RTX A5000 24 GB Q8_0 256K context · ~270 tok/s
- RTX PRO 4000 Blackwell 24 GB Q8_0 256K context · ~236 tok/s
- RTX Spark 24 GB Q8_0 256K context · ~105 tok/s
- Tesla P40 24 GB Q8_0 256K context · ~122 tok/s
- RTX 5090 32 GB Q8_0 256K context · ~629 tok/s
- RTX 5000 Ada 32 GB Q8_0 256K context · ~202 tok/s
- RTX PRO 4500 Blackwell 32 GB Q8_0 256K context · ~315 tok/s
- Tesla V100 32GB 32 GB Q8_0 256K context · ~316 tok/s
- L40S 48 GB Q8_0 256K context · ~303 tok/s
- RTX 6000 Ada 48 GB Q8_0 256K context · ~337 tok/s
- RTX A6000 48 GB Q8_0 256K context · ~270 tok/s
- RTX PRO 5000 Blackwell 48 GB Q8_0 256K context · ~472 tok/s
- RTX PRO 5000 Blackwell 72GB 72 GB Q8_0 256K context · ~472 tok/s
- RTX PRO 5500 Blackwell 84 GB Q8_0 256K context · ~491 tok/s
- RTX PRO 6000 Blackwell 96 GB Q8_0 256K context · ~629 tok/s
- DGX Spark (GB 10) 128 GB Q8_0 256K context · ~96 tok/s
AMD
- Radeon RX 5600 XT 6 GB Q8_0 256K context · ~101 tok/s
- Radeon RX 5700 8 GB Q8_0 256K context · ~157 tok/s
- Radeon RX 5700 XT 8 GB Q8_0 256K context · ~157 tok/s
- Radeon RX 6600 8 GB Q8_0 256K context · ~79 tok/s
- Radeon RX 6600 XT 8 GB Q8_0 256K context · ~90 tok/s
- Radeon RX 6650 XT 8 GB Q8_0 256K context · ~98 tok/s
- Radeon RX 7600 8 GB Q8_0 256K context · ~101 tok/s
- Radeon RX 9060 8 GB Q8_0 256K context · ~101 tok/s
- Radeon RX 9060 XT 8 GB Q8_0 256K context · ~112 tok/s
- Radeon RX 6700 XT 12 GB Q8_0 256K context · ~135 tok/s
- Radeon RX 6750 XT 12 GB Q8_0 256K context · ~152 tok/s
- Radeon RX 7700 XT 12 GB Q8_0 256K context · ~152 tok/s
- Radeon RX 9070 GRE 12 GB Q8_0 256K context · ~152 tok/s
- AMD BC-250 16 GB Q8_0 256K context · ~157 tok/s
- Radeon RX 6800 16 GB Q8_0 256K context · ~180 tok/s
- Radeon RX 6800 XT 16 GB Q8_0 256K context · ~180 tok/s
- Radeon RX 6900 XT 16 GB Q8_0 256K context · ~180 tok/s
- Radeon RX 6950 XT 16 GB Q8_0 256K context · ~202 tok/s
- Radeon RX 7600 XT 16 GB Q8_0 256K context · ~101 tok/s
- Radeon RX 7800 XT 16 GB Q8_0 256K context · ~219 tok/s
- Radeon RX 7900 GRE 16 GB Q8_0 256K context · ~202 tok/s
- Radeon RX 9060 XT 16GB 16 GB Q8_0 256K context · ~112 tok/s
- Radeon RX 9070 16 GB Q8_0 256K context · ~225 tok/s
- Radeon RX 9070 XT 16 GB Q8_0 256K context · ~225 tok/s
- Radeon RX 7900 XT 20 GB Q8_0 256K context · ~281 tok/s
- Radeon RX 7900 XTX 24 GB Q8_0 256K context · ~337 tok/s
- Radeon AI PRO R9700 32 GB Q8_0 256K context · ~225 tok/s
- Ryzen AI Max (Strix Halo) 32 GB Q8_0 256K context · ~90 tok/s
- Ryzen AI Max 400 (Gorgon Halo) 64 GB Q8_0 256K context · ~96 tok/s
Intel
- Arc A580 8 GB Q8_0 256K context · ~180 tok/s
- Arc A750 8 GB Q8_0 256K context · ~180 tok/s
- Arc A770 8GB 8 GB Q8_0 256K context · ~180 tok/s
- Arc B570 10 GB Q8_0 256K context · ~133 tok/s
- Arc B580 12 GB Q8_0 256K context · ~160 tok/s
- Arc A770 16GB 16 GB Q8_0 256K context · ~197 tok/s
- Arc Pro B65 32 GB Q8_0 256K context · ~214 tok/s
- Arc Pro B70 32 GB Q8_0 256K context · ~214 tok/s
Apple
- M1 8 GB Q8_0 256K context · ~24 tok/s
- M2 8 GB Q8_0 256K context · ~35 tok/s
- M3 8 GB Q8_0 256K context · ~35 tok/s
- M1 Pro 16 GB Q8_0 256K context · ~70 tok/s
- M2 Pro 16 GB Q8_0 256K context · ~70 tok/s
- M4 16 GB Q8_0 256K context · ~42 tok/s
- M5 16 GB Q8_0 256K context · ~54 tok/s
- M6 16 GB Q8_0 256K context · ~54 tok/s
- M3 Pro 18 GB Q8_0 256K context · ~53 tok/s
- M4 Pro 24 GB Q8_0 256K context · ~96 tok/s
- M5 Pro 24 GB Q8_0 256K context · ~108 tok/s
- M1 Max 32 GB Q8_0 256K context · ~141 tok/s
- M2 Max 32 GB Q8_0 256K context · ~141 tok/s
- M3 Max 36 GB Q8_0 256K context · ~105 tok/s
- M4 Max 36 GB Q8_0 256K context · ~144 tok/s
- M5 Max 36 GB Q8_0 256K context · ~162 tok/s
- M1 Ultra 64 GB Q8_0 256K context · ~197 tok/s
- M2 Ultra 64 GB Q8_0 256K context · ~197 tok/s
- M3 Ultra 96 GB Q8_0 256K context · ~201 tok/s
- M5 Ultra 96 GB Q8_0 256K context · ~295 tok/s