Pipette by Liquid AI

On-device model benchmark leaderboard

An open-source platform for the large-scale benchmarking of foundation models on devices.

Models benchmarked

  • LiquidAI/LFM2-2.6B-Exp-GGUF
  • LiquidAI/LFM2-2.6B-GGUF
  • LiquidAI/LFM2-700M-GGUF
  • LiquidAI/LFM2.5-1.2B-Instruct-GGUF
  • LiquidAI/LFM2.5-2.6B-GGUF
  • LiquidAI/LFM2.5-230M-GGUF
  • LiquidAI/LFM2.5-350M-GGUF
  • LiquidAI/LFM2.5-8B-A1B-GGUF
  • ibm-granite/granite-4.0-350m-GGUF
  • ibm-granite/granite-4.0-h-1b-GGUF
  • ibm-granite/granite-4.0-h-350m-GGUF
  • ibm-granite/granite-4.0-h-micro-GGUF
  • mistralai/Ministral-3-3B-Instruct-2512-GGUF
  • unsloth/Llama-3.2-1B-Instruct-GGUF
  • unsloth/Llama-3.2-3B-Instruct-GGUF
  • unsloth/Ministral-3-3B-Instruct-2512-GGUF
  • unsloth/Qwen3.5-0.8B-GGUF
  • unsloth/Qwen3.5-2B-GGUF
  • unsloth/Qwen3.5-4B-GGUF
  • unsloth/gemma-4-E2B-it-GGUF
  • unsloth/gemma-4-E4B-it-GGUF

Devices benchmarked

  • Galaxy S26 Ultra
  • MacBook Pro
  • Ryzen AI Max+ 395
  • iPhone 17 Pro