AI Developer Runs 28.9-Million-Parameter Model on $10 ESP32-S3 Microcontroller
Ukrainian developer Slava S (slvDev on GitHub) has successfully run a 28.9-million-parameter language model locally on an ESP32-S3 microcontroller, a chip costing under $10. The model, quantized to 4-bit (14.9 MB total), uses Google's Per-Layer Embeddings technique to overcome severe memory constraints. By storing the 25-million-parameter embedding table in the chip's 16MB Flash memory instead of the limited 512KB SRAM or 8MB PSRAM, the processor can efficiently access only needed rows per token. The model was trained on the TinyStories dataset and is limited to generating short fictional stories, not general chat or coding. The achievement is a proof-of-concept demonstrating that neural networks can run on extremely cheap embedded hardware, with potential applications like smart appliances (e.g., a coffee machine with expert knowledge) operating fully offline.
Editorial responsibility
- No named human review is recorded for this page.
- Reports are grouped by semantic similarity and deterministic rules. Language models may assist titles, summaries, translation and cross-source analysis; the page itself is projected from evidence records.
- Current automated evidence projection