← sources

GPT-6 Sol & Luna: Fool’s gold for the gullible

source: reddit (reddit:OpenAI)published 2026-09-23 17:21ingested 2026-09-24 02:38

retained text

GPT-6 Sol and Luna? More like fool’s gold dressed up in a shiny wrapper. These things feel dumber than 5.6, and no, a cheaper price tag doesn’t make that okay. My guess is OpenAI was trying to win back everyone who rage-quit over Astra eating through their tokens - and yeah, maybe take some heat off Astra too. But dumbing down your product and hoping nobody notices? Not a great look.

They threw out the budget models and people jumped at them - either to save cash or just to stop watching their subscription evaporate in real time. And honestly? It worked, at least for a bit. Astra was absolutely shredding through limits, so anything felt like a relief. But come on, OpenAI - this isn’t the move. Here’s hoping they start throwing in limit resets for every million users who bail.

document fetched 2026-09-24 00:30 · canonical https://reddit.com/r/OpenAI/comments/1woc9lt/gpt6_sol_luna_fools_gold_for_the_gullible