Unlocking ChatGPT's 1M Context Window: Costs Double After 272K Tokens
eyishazyer · x · 2026-08-17
Developer Tibo released the configuration to unlock the full 1M-token context window in ChatGPT, but highlighted a significant pricing trade-off.
- Pricing Threshold: Once input exceeds 272K tokens, the entire request is billed at 2x for input and 1.5x for output, not just the excess tokens.
- Availability: The feature works through regular ChatGPT accounts as well as API keys.
- Context: Such configurations typically remain undocumented until reverse-engineered, making this disclosure a genuine surprise.
More from Models
- Dev on OpenAI 1M context: Seamless compaction is the real win — eyishazyer · 2026-08-17
- Users report Qwen models "overthink" simple responses, exhausting context — malliktwts · 2026-08-17
- Local benchmark: Qwen 3.8 and DeepSeek 4 show significant quality jump on Mac — olcan · 2026-08-17
- Sarvam AI Unveils 105B Voice Model and Full-Stack Agent Solutions — AashaySachdeva · 2026-08-17
- GPT-4o Pro run time exceeds 2 hours in user test — banteg · 2026-08-17
- Local test: Qwen 3.8 27B's overthinking brings it near Sonnet-level performance — maxwell321 · 2026-08-17