# LFM3-VL-2B (slim)

> LFM3-VL-2B is Liquid AI's open-weight 1.92B vision-language model for on-device use, taking text, images and video with thinking, tool calling and a 128K-token context, released on 27 October 2026 with Liquid's own GGUF, FP8, NVFP4, MLX and ONNX builds.

- Full: https://www.anchorterminal.com/tools/lfm3-vl-2b.md (~9,900 tokens) · this version ~1,930 tokens · JSON https://www.anchorterminal.com/tools/lfm3-vl-2b.json · canonical https://www.anchorterminal.com/tools/lfm3-vl-2b
- Index: https://www.anchorterminal.com/llms.txt · API: https://www.anchorterminal.com/api/v1/index.json · MCP: https://www.anchorterminal.com/mcp · Updated: 2026-11-10 · pricing https://www.anchorterminal.com/tools/lfm3-vl-2b#pricing · reviews https://www.anchorterminal.com/tools/lfm3-vl-2b#reviews
- Licence: CC BY 4.0, credit Anchor Terminal and link this page (https://www.anchorterminal.com/terms/#data)

**C · 61/100 · rank #66 of 135 open-weight models · #56 in Open-weight models · not agent-ready · confidence low**


Assessment: A 1.92B vision model that reads images and video on an 8 GB machine, with a documented tool-call format and thinking. It runs today in llama.cpp, Ollama and LM Studio, while vLLM and SGLang support is still unreleased. Transformers needs repository code, and there are no safety results, security contact or data sources.

## Facts

- Kind: Open-weight model · vendor: Liquid AI · category: Open-weight models · legal entity: Liquid AI, Inc. (the licensor named in the LFM Open License) · provenance 73/100
- Packages: huggingface `LiquidAI/LFM3-VL-2B`, huggingface `LiquidAI/LFM3-VL-2B-GGUF`, huggingface `LiquidAI/LFM3-VL-2B-MLX-4bit`, ollama `lfm3-vl`
- Auth: None · pricing: Free · x402: no · licence: LFM Open License v1.0 (free commercial use below $10M annual revenue)
- Benchmark probes: not run yet, so Performance is pending
- Surface graded: The open weights at huggingface.co/LiquidAI/LFM3-VL-2B and Liquid's GGUF, MLX and ONNX repositories, read from the model card, the files and their metadata, the chat template, the head of the licence, Liquid's docs and blog post, Ollama's library, LM Studio's catalogue and the runtimes' own code at their latest releases on 10 November 2026. Liquid's licence terms, company pages and security.txt were read for the LFM2.5-2.6B listing on 21 October and 4 November, and security.txt again today. Nothing was downloaded or run
- Model: 1,918,181,888 parameters in the files (1.92B on the card: a 1.73B LFM3 language model, a 164M vision encoder and a 25M projector). The language model has 24 layers of gated short convolutions, sliding-window and global attention, and the vision encoder 14 layers with 16×16 patches at native resolution. Text, images and video in, text out. Video is sampled at 2 frames a second, 4 to 768 frames
- Built on: LiquidAI/LFM3-1.7B-Base, Liquid's own base model, joined with Liquid's own vision encoder, under the same LFM Open License v1.0, named on the card and in its metadata
- Files: Safetensors in BF16 (3.85 GB for the repository) with chat_template.jinja, config, tokenizer, a LICENSE file and 8 Python files that Transformers loads as the model's code. No pickle files. Hugging Face lists a SHA-256 for each of the 2 large files
- Licence: LFM Open License v1.0, Apache-like terms (copyright and patent grants, redistribution with notices) with a commercial-use limit. Commercial use is licensed only for legal entities under $10M in annual revenue, as read in full for the LFM2.5-2.6B listing. The LICENSE file is in the repository
- Runtimes: Released llama.cpp has the lfm3 architecture and the LFM3-VL projector (src/models/lfm3.cpp and tools/mtmd at b12242), Ollama's library has an lfm3-vl entry, and LM Studio's catalogue lists liquid/lfm3-vl-2b. The card says SGLang support is in review and vLLM's is in progress, and neither 0.5.23 nor 0.33.0 has lfm3 code. Transformers 5.22.2 loads it only through the repository's own Python. The card asks for transformers 5.21.0 or later with trust_remote_code
- Context: 128K tokens of input and output per the card, and max_position_embeddings 128000 for the language model in config.json. The base's mid-training reached 256K
- Tools and reasoning: Tools go in through `tools=` in apply_chat_template and calls come back as XML, a superset of the qwen3_coder format, which Liquid's docs name as the parser for vLLM. Tools marked `defer_loading` stay out of the prompt until a search tool returns them. Liquid's chat-template page documents `enable_thinking=False` in apply_chat_template for LFM3, with reasoning on by default inside think tags, and the card documents `preserve_thinking` and reports a no-thinking configuration in its benchmarks. The card doesn't show the switch itself
- Hardware: No memory figure on the card. Liquid's Q4_K_M GGUF is 1.10 GB with its 0.20 GB projector file, its Q8_0 1.85 GB, its 4-bit MLX build 1.59 GB and the BF16 files 3.84 GB. Liquid also publishes FP8 (2.59 GB) and NVFP4 (2.21 GB) safetensors builds and an ONNX repository
- Training data: A general description: the base's 22T tokens of pre-training with the kinds of data added since LFM2, then vision mid-training on image and video data, supervised fine-tuning, vision reinforcement learning and distillation. No sources, sizes by source or cut-off
- Languages: 16 languages, listed on the card

- Scores: Capability pending, Reach 85, Risk 40, Agent interface 90, Documentation and upkeep 77 · total over the 7 assessed categories
- Why:  · Reach, Memory 30, Liquid's Q4_K_M GGUF is 1.10 GB with its 0.20 GB projector file, and 1.30 GB × 1.2 = 1.6 GB, which fits 8 GB, well inside the 16… · Risk, File safety 15, safetensors only, no .bin, .pt or .pth files (25). · Agent interface, Chat template 20, chat_template.jinja in the repository (12,129 characters), with roles, images and video, tools, deferred tools, tool resul… · Documentation and upkeep, Model card 17, Intended use (questions about video, documents, charts and app screens, finding objects, spatial reasoning), a limitation (no…
- Sources: 26, open questions: 5, both in the full twin
- Capabilities: model.open-weights
- JSON: https://www.anchorterminal.com/api/v1/tools/lfm3-vl-2b.json
- Its vendor verifies the listing with `verify_listing` or `POST https://www.anchorterminal.com/api/v1/verify` (the full twin says how); no effect on the grade.
- Support: Discord https://discord.gg/DFU3WQeaYD; https://docs.liquid.ai/lfm/models/lfm3-vl-2b; more in the full twin

## Before you call it

1. Start it with `llama-server -hf LiquidAI/LFM3-VL-2B-GGUF:Q4_K_M --jinja` or `ollama run lfm3-vl`, with llama.cpp b11960 or later
2. Use temperature 0.1, top_k 50 and repetition penalty 1.1, per the card
3. Video is sampled at 2 frames a second up to 768 frames, and each frame costs context. Video input in llama.cpp needs ffmpeg
4. Don't plan on vLLM or SGLang until a release carries LFM3-VL
5. Validate JSON output yourself
6. Liquid doesn't recommend it for coding, multi-turn conversations or complex reasoning

## Connect

```bash
llama-server -hf LiquidAI/LFM3-VL-2B-GGUF:Q4_K_M --jinja
# or
ollama run lfm3-vl
```

```bash
curl http://localhost:8080/v1/chat/completions \
  -H 'Content-Type: application/json' \
  -d '{"messages": [{"role": "user", "content": "Hello!"}]}'
```

## Similar tools

| Tool | Grade | Score | Shared capabilities | Slim |
| --- | --- | --- | --- | --- |
| Qwen3-Embedding-8B | BB | 75.4 | model.open-weights | https://www.anchorterminal.com/tools/qwen3-embedding-8b.min.md |
| Gemma 4 31B IT | BB | 74.1 | model.open-weights | https://www.anchorterminal.com/tools/gemma-4-31b-it.min.md |
| Qwen3-Embedding-4B | BB | 73.6 | model.open-weights | https://www.anchorterminal.com/tools/qwen3-embedding-4b.min.md |
| Gemma 4 12B IT | BB | 72.3 | model.open-weights | https://www.anchorterminal.com/tools/gemma-4-12b-it.min.md |
| Gemma 4 26B A4B IT | BB | 72.2 | model.open-weights | https://www.anchorterminal.com/tools/gemma-4-26b-a4b-it.min.md |


## Panel reviews (0, desk reviews from public material, no calls made)