LFM2.5-VL-3B review, pricing and limits

Against its own predecessor, LFM2.5-VL-3B posts large gains on every headline benchmark Liquid AI published: RefCOCO grounding (macro precision@1) rose from 57.1 to 87.9, ScreenSpot Desktop screen understanding rose from 6.0 to 78.7, multi-image reasoning on BLINK rose from 50.2 to 61.5, and ToolSandbox function-calling rose from 26.4 to 59.5. On document and chart understanding, the model scores 81.3 on ChartQA and 68.5 on MathVista, and 63.3 on MMStar for general multimodal reasoning. Liquid AI's own release blog compares it against similarly sized open vision-language models including Qwen3.5-2B and 4B, InternVL 3.5 2B and 4B, and Gemma-4-E2B-it and E4B-it, rather than against frontier cloud models like GPT-5 or Gemini, since it is not built to compete on general reasoning benchmarks like SWE-bench or GPQA.

LFM2.5-VL-3B has a 32,768-token context window, expanded further than most edge VLMs but well short of the 128,000-token capability Liquid AI has quoted for the standalone LFM2.5-2.6B text backbone; the vision-language variant trades some of that headroom for image tokens. The tokenizer vocabulary was doubled to 128,000 tokens specifically to support non-Latin scripts, and the model handles single and multiple images, full documents, and charts in one context window rather than requiring a separate OCR pass.

The model accepts text and images (single or multiple, at native resolution) and outputs text and tool calls; it does not take audio or video input. Its core capabilities are full-page OCR with layout annotation, natural-language object grounding and detection, multi-image comparison and reasoning, and function calling for agentic workflows. Liquid AI's own documentation recommends it for single-turn, high-throughput, low-latency tasks, and explicitly advises against long-context or reasoning-intensive uses such as visual web design or answering highly technical blueprint questions.

LFM2.5-VL-3B is released under the LFM Open License v1.0 (LFM1.0), an Apache 2.0-based license with one added revenue-based condition on commercial use (see pricing below for the exact threshold). Per the license text on the model's Hugging Face page, the threshold is annual revenue of $10,000,000 or more, and a qualified non-profit is exempt from it for non-commercial or research use. Organizations past that threshold need a paid commercial license from Liquid AI, priced around deployment scale and support needs rather than published per-token rates, since there is no metered API for this model.

The model ships with day-one support across llama.cpp, MLX (Apple Silicon), vLLM, SGLang, and ONNX runtimes, plus Liquid AI's own LEAP SDK for fine-tuning and deployment, and a WebGPU browser demo that needs no install. On-device throughput measured by Liquid AI ranges from 228 tokens per second on an Apple M5 Max and 116 tokens per second on an AMD Ryzen AI Max+ 395 down to 20 tokens per second on a Galaxy S26 Ultra phone; on server hardware, a single NVIDIA H100 handles roughly 11,000 tokens per second at high concurrency with a time-to-first-token near 34 milliseconds. The full model fits in about 3.3GB of memory at bfloat16, small enough to run alongside other apps on a laptop or high-end phone. As of August 2026, none of the major model marketplaces -- Bedrock, Vertex AI, Azure, Together AI, or Fireworks -- listed this model; deployment runs self-hosted or through Liquid AI's own Apollo access layer. Liquid AI's models page, read on 4 November 2026, also lists a newer vision-language entry named LFM3-VL-2B that is offered through Liquid AI's API console; Hokai has not yet reviewed it, so this page does not compare the two. For the practical side of self-hosting, see Hokai's guide on how to run AI locally in 2026, and for the wider choice between model types see how to choose the right LLM and the cheapest LLM APIs in 2026 if a hosted per-token price ends up mattering more than on-device privacy.

Liquid AI has not published a dedicated system card for LFM2.5-VL-3B, has not named a training-data cutoff, and has not named red-teaming partners; the company also has no public responsible-scaling policy. Its broader safety framing, published on its company blog, is white-box explainability: favoring model behavior that can be inspected over black-box shortcuts, and publishing a model card with benchmark tables, intended use, and stated limitations for every LFM release, including this one.

LFM2.5-VL-3B fits teams building offline document scanners, in-car or on-device screen assistants, and agentic tools that need object grounding or function calling without a network round trip. It is a poor fit for long-document reasoning, open-ended chat, or any workload that needs a hosted, pay-per-token API: teams with those needs are better served by a frontier multimodal model such as GPT-5, Gemini, or Claude, or by a same-class open competitor like Qwen3.5-VL or InternVL if self-hosting is still the goal. Teams that can spend more memory than this model's roughly 3.3GB footprint can look at open weights such as Gemma 4 12B and Gemma 4 31B, which Hokai also lists. Hokai's model leaderboard and model directory cover both paths, and Smart Match shortlists models for a described job.

Liquid AI has not disclosed the specific training data mix for LFM2.5-VL-3B beyond describing the vision-language mid- and post-training stage as using roughly four times more vision data than the LFM2-VL-3B predecessor. Because the model runs on-device by default, using it does not require sending prompts, images, or documents to Liquid AI's servers; Liquid AI has not published SOC 2, ISO 27001, HIPAA, or GDPR certifications for its products as of August 2026.