memovai/mimimodel

Running a LLM on the ESP32

Automation foundation model for tiny devices: 2-bit, 8-29 MB, tool calls, ASR, structured extraction and embeddings on phones, wearables, smart homes, robots, cars and microcontrollers.

Distributed 56M-parameter LLM inference across 3 ESP32-S3 boards via ESP-NOW , Split-PLE + KV cache, fully offline.

Run a 1-billion parameter LLM on a $10 board with 256MB RAM

Fully offline tiny LLM chatbot on the ESP32-S3 Cardputer ADV - 8M params, Q4 weights baked into the firmware, ~5 tok/s.

On-device LLM Inference Powered by X-Bit Quantization

MimiClaw: Personal Agent on a $5 chip. No OS(Linux). No Node.js. No Mac mini. No Raspberry Pi. No VPS. Hardware agents OS. On device Muse.

Quantization, kernels, runtime and inference engine for mobiles, wearables, smart home and robots.

Local AI API Platform