DeepSeek V4 Flash Sets ARC-AGI Cost-Performance RecordARC Prize Foundation verifies the model's benchmark scores at low per-task cost.Aug 7, 2026 · 18 posts
Greg Kamradt Runs 78K Classifications on GPT-5.6 Luna for $60AI educator processed 158K requests and 143M tokens overnight using batch mode.Jul 31, 2026 · 4 posts