Model: LFM3-VL 2B

Liquid AI released LFM3-VL 2B, the vision-language member of its LFM3 family, pairing a 24-layer text model shaped like LFM3 1.7B with a 14-layer vision encoder shipped as a separate 379 MB F16 projector file. The model carries a 128,000-token text window — half of LFM3 1.7B's — and ships in four GGUF quantization variants on Hugging Face, from Q4_K_M at 1.10 GB to Q8_0 at 1.85 GB, each requiring the 0.38 GB projector as a separate download; llama.cpp and Ollama both run it. Only 6 of the 24 text layers keep a cache that grows with context, while 5 use a 256-token sliding window and 13 are short convolutions.

LFM3-VL 2B 2B Liquid AI LFM3 The vision-language member of Liquid AI's LFM3 family, with the same 24-layer text model shape as LFM3 1.7B. A 14-layer vision encoder ships as a separate projector file 379 MB at F16, loaded alongside the model . Only 6 of the 24 text layers keep a cache that grows with context; 5 use a 256-token sliding window and 13 are short convolutions. Its text window is 128,000 tokens, half of LFM3 1.7B's. llama.cpp and Ollama both run it. 128K Max Context 4 Quant Variants GGUF Q8 0 Best Quality — Perplexity rise best build Quantization Variants Per-quant VRAM, quality loss, and inference speed on RTX 4090 Measured = run on this site · Estimated = not run here: calculated, or a figure this site has not verified · Community = public reports | Format | Level | BPW | VRAM | PPL Loss | Speed | Source | Actions | |---|---|---|---|---|---|---|---| | GGUF | Q4 K M | 5.08 | 1.2 GB | — | — | Estimated | | | GGUF | Q5 K M | 5.82 | 1.4 GB | — | — | Estimated | | | GGUF | Q6 K | 6.6 | 1.6 GB | — | — | Estimated | | | GGUF | Q8 0 | 8.54 | 2.0 GB | — | — | Estimated | | Which file each size is for A file size is not a memory figure: the VRAM column is the calculator's total at 4K context — weights plus KV cache plus a 10% activation buffer. Units differ on purpose: file sizes are decimal GB, as Hugging Face lists them; the VRAM column is GiB 1024³ bytes with 2% added to the weights, so 7.12 GB of file counts as about 6.8 GiB there. A vision projector mmproj , where the repo ships one, is a separate download and is in neither number. - GGUF Q4 K M LiquidAI/LFM3-VL-2B-GGUF https://huggingface.co/LiquidAI/LFM3-VL-2B-GGUF · LFM3-VL-2B-Q4 K M.gguf · 1.10 GB 1,099,627,808 bytes · + vision projector 0.38 GB, separate · listing read 2026-11-03 - GGUF Q5 K M LiquidAI/LFM3-VL-2B-GGUF https://huggingface.co/LiquidAI/LFM3-VL-2B-GGUF · LFM3-VL-2B-Q5 K M.gguf · 1.26 GB 1,258,663,200 bytes · + vision projector 0.38 GB, separate · listing read 2026-11-03 - GGUF Q6 K LiquidAI/LFM3-VL-2B-GGUF https://huggingface.co/LiquidAI/LFM3-VL-2B-GGUF · LFM3-VL-2B-Q6 K.gguf · 1.43 GB 1,427,638,304 bytes · + vision projector 0.38 GB, separate · listing read 2026-11-03 - GGUF Q8 0 LiquidAI/LFM3-VL-2B-GGUF https://huggingface.co/LiquidAI/LFM3-VL-2B-GGUF · LFM3-VL-2B-Q8 0.gguf · 1.85 GB 1,846,238,496 bytes · + vision projector 0.38 GB, separate · listing read 2026-11-03 Similar models Compare with LFM3 1.7B https://quantized.uk/tools/compare/?a=lfm3-vl-2b&b=lfm3-1-7b 1.7BConsumer GPUMac / Apple SiliconQ4 K M1.2 GB—perplexity rise https://quantized.uk/quant-hub/lfm3-1-7b/ LFM3 1.7B Liquid AI LFM3 The smallest model of Liquid AI's LFM3 family, released on 27 October: a 1.7B text model with a 256K-token window, built for phones, laptops and small GPUs. Only 6 of its 24 layers keep a cache that grows with context; 5 use a 256-token sliding window and the other 13 are short convolutions. At Q4 K M it needs about 1.2 GB at 4K and 2.8 GB at 128K, comfortable on an 8 GB card; the full window needs about 4.5 GB at 256K, comfortable on an 8 GB card. Q8 0 needs about 2.0 GB at 4K. Ollama runs the LFM3 family on Windows, Linux and Apple silicon since v0.40.13, and llama.cpp loads its architecture on every platform. Sizes come from Liquid's own GGUF files, and no per-level quality loss has been published. 2BConsumer GPUMac / Apple SiliconQ4 K M1.5 GB—perplexity rise https://quantized.uk/quant-hub/qwen3-5-2b/ Qwen3.5 2B Alibaba Qwen3.5 The 2B size of Qwen3.5, small enough for almost anything: old cards, base-model Macs, mini PCs and CPU-only servers. Only 6 of its 24 layers keep a growing KV cache, with 2 KV heads each. At Q4 K M it needs about 1.5 GB at 4K context, 1.9 GB at 32K and 4.8 GB at its full 256K window. Even Q8 0 stays around 2.2 GB at 4K. Reasoning is off by default on the small Qwen3.5 sizes. Sizes come from the published GGUF file sizes, and no per-level quality loss has been published. 3BConsumer GPUMac / Apple SiliconQ4 K M2.5 GB+3.5%perplexity rise https://quantized.uk/quant-hub/llama-3.2-3b/ Llama 3.2 3B Instruct Meta Llama 3.2 Tiny but capable. Runs on 4GB VRAM or 8GB RAM, even on phones via llama.cpp. 3BConsumer GPUMac / Apple SiliconQ4 K M2.1 GB+3.2%perplexity rise https://quantized.uk/quant-hub/qwen2.5-3b/ Qwen2.5 3B Instruct Alibaba Qwen2.5 Tiny Qwen2.5 for edge devices. Runs on 4GB VRAM or Raspberry Pi class hardware. Running LFM3-VL 2B locally At Q4 K M and 4K of context, LFM3-VL 2B needs about 1.2 GB — 1.0 GB of weights, 0.05 GB of KV cache and a 0.1 GB activation buffer. The smallest cards in this index that clear that comfortably are 8 GB ones — 11 of them, from the Mac M1 8G up — and 78 of the 78 cards here do. These are calculated figures, not measurements: the estimate stops counting a card as comfortable at 88% of its VRAM, which is roughly the room a desktop session needs. What longer context costs Going from 4K to 32K adds about 0.33 GB, taking the total to 1.6 GB. That is gentler than the parameter count suggests: only 6 of 24 layers keep a KV cache that grows with context, and 5 more cap theirs at a sliding window, so the cache scales at a fraction of the usual rate. The model's native window is 125K; holding all of it at Q4 K M would need about 3 GB. Which build to download This index only tracks LFM3-VL 2B in GGUF, across 4 levels Q4 K M, Q5 K M, Q6 K, Q8 0 . No per-level perplexity sweep has been published for this model, so the quality column is empty rather than estimated — within one model, more bits per weight is the only ordering the data supports. No throughput has been measured for this model on this site's hardware. GGUF runs on llama.cpp and Ollama across NVIDIA, AMD and Apple silicon; AWQ and GPTQ target vLLM on CUDA and ROCm; EXL2 is ExLlamaV2 — CUDA only, and now archived, so local use rather than serving. Common questions - How much VRAM does LFM3-VL 2B need? - About 1.2 GB at Q4 K M with 4K of context and batch 1, rising to roughly 1.6 GB at 32K. That figure is weights plus KV cache plus a 10% activation buffer, calculated from the model's architecture rather than measured on a card. - Will LFM3-VL 2B run on a 8GB GPU? - Yes — at Q4 K M and 4K context it needs about 1.2 GB, which leaves 4.8 GB spare on a Mac M1 8G, or on any of the other 8 GB cards here 11 in all . That is the smallest size in this index that clears it comfortably; 78 of 78 do. - Which quantization of LFM3-VL 2B should I use? - No published quality comparison exists for this model, so pick by footprint: Q4 K M is the level most people run, and a higher bits-per-weight level is more faithful. All 4 levels in the index are Q4 K M, Q5 K M, Q6 K, Q8 0. Where this model fits Sized at Q4 K M with a 4K context window, smallest card first. Comfortable means the estimate uses at most 88% of the memory. - Runs comfortably on - Mac M1 8G8GB · +4.8 https://quantized.uk/gpu/mac-m1-8g/ Mac M3 8G8GB · +4.8 https://quantized.uk/gpu/mac-m3-8g/ RTX 40608GB · +6.8 https://quantized.uk/gpu/rtx-4060/ RTX 4060 Ti 8G8GB · +6.8 https://quantized.uk/gpu/rtx-4060-ti-8g/ Radeon RX 9060 XT 8G8GB · +6.8 https://quantized.uk/gpu/radeon-rx-9060-xt-8g/ - Ships in - GGUF https://quantized.uk/formats/gguf/ - Try it yourself - Size it in the calculator https://quantized.uk/tools/vram-calc/?model=lfm3-vl-2b&quant=Q4 K M Generate the run command https://quantized.uk/tools/cli-gen/?model=lfm3-vl-2b&quant=Q4 K M How to actually run this Deployment guides for this model and this class of hardware. This page's figures change when the model or the runtime does.Last updated 2026-11-03 RSS → /feed.xml https://quantized.uk/feed.xml