devtake.dev — #on-device-aiArticles tagged on-device-ai on devtake.dev.https://devtake.dev/en-usCactus Compute distilled Gemini into a 26M tool-calling model. The trick: no feed-forward layers.https://devtake.dev/article/needle-cactus-compute-tool-calling/https://devtake.dev/article/needle-cactus-compute-tool-calling/Needle is a 26M-parameter function caller distilled from Gemini 3.1 Flash-Lite. The Simple Attention Network drops MLPs and runs at 6,000 tok/s prefill on edge silicon.Wed, 13 May 2026 10:00:00 GMTaiai-modelsgeminiopen-weightsfunction-callingtool-callingedge-aion-device-aismall-modelsdieter-morelliAnker built its own AI chip. It runs neural nets inside flash memory cells.https://devtake.dev/article/anker-thus-on-device-ai-chip/https://devtake.dev/article/anker-thus-on-device-ai-chip/Anker's Thus chip embeds compute inside NOR flash, claims 150x more on-device AI for noise cancellation, and ships in Soundcore earbuds on May 21. Here's why it matters.Mon, 27 Apr 2026 08:00:00 GMThardwareankerai-chipscustom-siliconon-device-aicompute-in-memorysoundcoresemiconductorhiro-tanaka