Nous' Teknium Warns Against GPT 900k Variant Over Extra Billing
Nous Research's Teknium warned users to avoid the 900k-context GPT variant because usage beyond 272k context triggers extra billing, while a heavy user reported burning 19.7 billion tokens in 19 days even on the Ultra plan, and Nous teased upcoming caching improvements.
2026-11-16 ~ 2026-11-16 · 3 related posts
- Power user burns 19.7B tokens on Hermes — even Ultra can't cover a week — HeyJuano · 2026-11-16
- Teknium: avoid 900k GPT variants — they charge extra beyond 272k context — Teknium · 2026-11-16
- Nous exec: avoid GPT 900k context variant — extra charges kick in past 272k — Teknium · 2026-11-16