Run Gemma 4 26B in ~2 GB RAM on M-series Macs
Open-source local inference engine that compresses and runs large language models on Apple Silicon with very low memory. Suited to developers who want offline Gemma-class models on Mac without a discrete GPU.
| Resource | Pricing | Type | Licence | Platforms |
|---|---|---|---|---|
| Turbo Fieldfare | Open source | Tool | Apache-2.0 | mac |
| Kitten TTS | Open source | Tool | Apache-2.0 | web |
| AI Goofish Monitor | Open source | Tool | MIT | web |
| Steel Browser | Open source | Tool | Apache-2.0 | web · api |
| Open Design | Open source | Tool | Apache-2.0 | web |
| Airllm | Open source | Tool | Apache-2.0 | web |
| Open Generative AI | Open source | Tool | MIT | self-hosted |
25MB CPU-Only, Open-Source TTS Model
A Goofish multi-task real-time and scheduled monitoring and intelligent analysis system built with Playwright and AI, with a full-featured admin UI. It helps users find products they want among the large number of Goofish listings.
🔥 Open Source Browser API for AI Agents & Apps.
Local-first, open-source alternative to Claude Design.
AirLLM 70B inference with single 4GB GPU
Generate images and videos with 200+ models — Flux, Kling, Veo, Seedance and more. No account required.
