Liquid AI
About Liquid AI
Liquid AI develops foundation models with a stated focus on execution outside data centers. Its public documentation describes Liquid Foundation Models and deployment resources for multiple local runtimes. The company's device-oriented emphasis provides context for observing model behavior under explicit hardware and execution conditions.
Focus
- Device-oriented foundation models
- Local-runtime integrations
- Language and multimodal model variants
Projects and interfaces
Liquid Foundation Models
model family
Published as LFM releases with model-specific tasks and deployment resources.
Runtime integration resources
deployment documentation
Documentation links include llama.cpp, MLX, ONNX, Ollama, SGLang and vLLM routes.
Scope and limitations
- Device support, performance and data handling depend on the selected model, runtime and configuration.
Official references
- Liquid AI official website. Checked .
Models
76 named catalog records. These are name and distribution records, not a count of distinct model architectures.
- d1
- d1-omni
- d1-omni:600m
- d1-omni:600m-bf16
- d1-omni:600m-q8_0
- d1-omni:latest
- d1:3b
- d1:3b-q8_0
- d1:latest
- jewelzufo/LFM2.5-350M-GGUF:latest
- jpmarindiaz/lfm2.5-vl-450m:latest
- lfm2
- lfm2:24b
- lfm2:24b-a2b
- lfm2:24b-bf16
- lfm2:24b-q4_K_M
- lfm2:24b-q8_0
- lfm2:latest
- lfm2.5
- LFM2.5-1.2B-Instruct
- LFM2.5-1.2B-Thinking
- lfm2.5-thinking
- lfm2.5-thinking:1.2b
- lfm2.5-thinking:1.2b-bf16
- lfm2.5-thinking:1.2b-q4_K_M
- lfm2.5-thinking:1.2b-q8_0
- lfm2.5-thinking:latest
- lfm2.5:24b
- lfm2.5:24b-a2b-bf16
- lfm2.5:24b-a2b-q4_0
- lfm2.5:24b-a2b-q4_K_M
- lfm2.5:24b-a2b-q8_0
- lfm2.5:8b
- lfm2.5:8b-a1b-bf16
- lfm2.5:8b-a1b-q4_K_M
- lfm2.5:8b-a1b-q8_0
- lfm2.5:latest
- lfm3
- lfm3-audio
- lfm3-audio:2b
- lfm3-audio:2b-mlx-bf16
- lfm3-audio:2b-mxfp8
- lfm3-audio:2b-nvfp4
- lfm3-audio:latest
- lfm3-omni
- lfm3-omni:2b
- lfm3-omni:2b-mlx-bf16
- lfm3-omni:2b-mxfp8
- lfm3-omni:2b-nvfp4
- lfm3-omni:latest
- lfm3-vl
- lfm3-vl:2b
- lfm3-vl:2b-mlx-bf16
- lfm3-vl:2b-mxfp8
- lfm3-vl:2b-nvfp4
- lfm3-vl:latest
- lfm3:1.7b
- lfm3:1.7b-mlx-bf16
- lfm3:1.7b-mxfp8
- lfm3:1.7b-nvfp4
- lfm3:latest
- liquid/lfm-2.5-2.6b
- liquid/lfm-2.5-2.6b:free
- liquid/lfm-2.5-24b-a2b
- LiquidAI/lfm2.5-1.2b-instruct:latest
- LiquidAI/lfm2.5-350m:latest
- oamazonasgabriel/lfm2-1.2b-tool:f16-8gbGPU
- oamazonasgabriel/lfm2.5-2.6b:q4_k_m-8gbGPU
- oamazonasgabriel/lfm2.5-230m:bf16-8gbRAM
- oamazonasgabriel/lfm2.5-8b-a1b:q4_k_m-8gbGPU
- sam860/LFM2:1.2b
- sam860/LFM2:2.6b
- sam860/LFM2:350m
- sam860/LFM2:700m
- sam860/LFM2:8b
- scorpion7slayer/LFM2.5-2.6B-GGUF:Q6_K
Model families
- lfm2.5-thinking
- Liquid AI (catalog grouping)