48-hour test: GPT-6 Sol on $200 plan and Opus 5.5 on $20 each burn only ~8% weekly limits
i_dg23 · x · 2026-09-25
A developer shared usage numbers after nearly 48 hours of heavy use of GPT-6 Sol and Claude Opus 5.5:
- GPT-6 Sol (High tier, no Fast Mode, $200 plan): a full day of heavy use consumed only 8% of the weekly limit
- Claude Opus 5.5 (Mid tier, $20 plan): also 8% of the weekly limit, while producing 4-5 website drafts for potential clients
The author praises OpenAI's work on Sol and notes Anthropic improved value via limit resets and higher usage caps. His workflow: let the two models collaborate — one builds, the other challenges and improves the output.
More from Models
- Micius-Flash: 27B Multimodal Model for Hardware and Semiconductor Agents — zhiman-ai · 2026-11-11
- Baidu says RSI auto-optimized ERNIE inference, 6.24x decode speedup on Kunxin P800 — xiaohu · 2026-11-11
- Baidu's ECHO: multi-token prediction reusing the last layer as a drafter — xiaohu · 2026-11-11
- ERNIE 5.5 Flash Preview: 280B-param multi-router MoE activating only 14B per token — xiaohu · 2026-11-11
- ERNIE 5.5 recreates the Pelican test: 117-second animation drawn frame-by-frame in Python — xiaohu · 2026-11-11
- 35,000+ Grok-built apps exposed patient data, plaintext passwords, and sales pipeline — The Decoder · 2026-11-11