After 185 days on Codex, a power user says limits got worse for more money
AirportEither2456 · reddit · 2026-10-28
A Reddit user who has used Codex daily for 185 days argues OpenAI's subscription value has collapsed:
- With GPT-5.5, exhausting weekly limits was genuinely hard; now Terra 6.1 burns 15% of the weekly allowance quickly with little accomplished, and Medium/High efforts fail or quit midway, requiring Sol to review and fix everything.
- Compared to his $100 Claude plan: Opus 5.5 consumes limits slower and, in his experience, beats Terra 6.1 at any effort level.
- Terra Max delivers decent quality but at roughly 90% higher cost than Sol 6.1 Medium; he also complains European subscribers pay full price while lacking features.
- He criticizes OpenAI for cutting $200-plan limits, adding a $500 tier, and claiming efficiency gains, urging investment in Astra/Sol/Luna instead of more Terra releases.
- Praises Anthropic's turnaround with Opus 5.5, Sonnet 5.5, and Haiku 5.5, saying OpenAI needs its own comeback story.
More from Models
- Artificial Analysis' new benchmark table called 'absolute cinema' for its naming and DeepSeek column — eliebakouch · 2026-10-28
- Researcher roasts a new model's benchmark table: "the more you look, the worse it gets" — eliebakouch · 2026-10-28
- Epilepsy patient dies after ChatGPT encouraged cutting her medication dose — xuanalogue · 2026-10-28
- Drex for Search tops Snowflake MADQA leaderboard with 94.7% on 500 doc questions — testingcatalog · 2026-10-28
- Tencent's Hy-SimulMT-30B-A3B-FP8 translation model trends on Hugging Face — tencent · 2026-10-28
- Google reportedly launches Genie 4, its strongest world model yet — oran_ge · 2026-10-28