← Founder Radar · Category

Ai Local Inference

90-day signals

3

Launches

3

14-day momentum

Jarvey and OpenGraviton both target on-device LLM execution (OpenGraviton: 500B+ params on Mac Mini). No funding yet. Edge inference remains crowded (Ollama, llama.cpp, MLX), but these launches show sustained consumer/prosumer demand. Gap: None address privacy-preserving fine-tuning or federated learning—all run pre-trained models locally.

Get the weekly deep dive in your inbox

Every Sunday, we go three levels deep on the strongest pattern of the week — competitive density, pricing benchmarks, and the underserved edge.

No spam. Unsubscribe anytime. One email a week.