Power user trusts Claude with open-ended budgets, says usage could 10x easily
AaronBergman18 · x · 2026-09-06
Aaron Bergman says a single $200 Claude plan once seemed massive, but after finding the model uniquely good and trustworthy, he now hands off almost any computer-based task with task budgets set only by API key permissions and account balance. He believes he could easily 10x (maybe 100x) his usage, and reveals he runs 4x $200 plans, using about $250/$800 for personal work.
More from Models
- Anthropic's real problem: compute cost plus low limits make beating GPT-6 Astra a must — haider1 · 2026-09-06
- Four top labs ship in one week: GPT-6 Astra, Gemini 3.8 Flash, Muse Spark and Claude updates — testingcatalog · 2026-09-06
- Fable 5.1 answers medical questions without refusals or Opus handoffs — JeremyNguyenPhD · 2026-09-06
- repligate: Anthropic's belief it can control Claude personalities is 'almost delusional' — repligate · 2026-09-06
- Unverified: third-party Scry claims 71.8% vs Gemini's 66.1% on DeepSearchQA — AaronBergman18 · 2026-09-06
- GPT-6 Sol reportedly in internal testing at OpenAI, possible release within weeks — koltregaskes · 2026-09-06