unfortunately, fully local models aren't the answer for people because these people aren't technical. they aren't going to run powerful computers with powerful GPU's costing thousands of dollars and aren't going to use llama and other open source models. i recommended nostr:npub10hpcheepez0fl5uz6yj...