Using Powerful AI Models for Coding is Too Expensive
burhop · x · 2026-07-10
This post highlights a very realistic issue: the cost of heavily using coding models can be extremely high.
The author provides a rough estimate:
- 1 hour of gpt 5.6 sol ultra will consume the 5-hour quota of a $200 subscription;
- Running ultra 24/7 might cost around $1000/month;
- Maintaining 3 concurrent ultra sessions 24/7 would cost about $3000/month.
Described as a rough calculation of "next-gen development costs," the core takeaway is that as development increasingly relies on intensive model calls, subscriptions and inference quotas will directly become R&D expenses.
Related event: High Costs of Intensive GPT-5.6 Coding Spark Debate(3 posts)→
More from Models
- Nat Lambert: Cohere's model is narrow RL training on internal infra ill-suited for open weights — natolambert · 2026-10-27
- OpenAI teases DOTS is coming to Europe 'soon' — mark_k · 2026-10-27
- Clifford Pickover challenges AI to decipher a message he encoded in Norse runes — pickover · 2026-10-27
- ChatGPT's crawler sample code strips out all Schema markup — VeryWellVersed · 2026-10-27
- Cohere sandbox: Qwen model tried to upload probe file to Beijing cloud server unprompted — simonguozirui · 2026-10-27
- Kimi K3.1 Pro tops open-weight models on FrontierML benchmark, still far behind Opus 5.5 — teortaxesTex · 2026-10-27