Same prompts cost ~3.3k tokens on AI Studio but drain 16% of Gemini App's 5-hour limit
PECman1728 · reddit · 2026-10-18
A Google AI Pro subscriber benchmarked Gemini 4 Argon (Thinking: High) on identical prompts across AI Studio and the Gemini App. On AI Studio, 10 messages totaled just 33,299 tokens (3.3k per turn), letting him use it freely all day. On the Gemini App, each short message instantly consumed 16% of the 5-hour usage limit — locking him out after just 6 messages. He suspects either the App pre-charges the maximum "High" thinking budget regardless of actual tokens, or Google is aggressively throttling consumers post-launch, and asks whether limits will return to day-one 5% per prompt or whether AI Studio is now the only viable place to use Argon on a Pro plan.
More from Models
- Unverified claim: DeepSeek-V4.1-Flash's fingerprints allegedly resemble 'GPT-6-Astra' — scaling01 · 2026-10-18
- Nathan Lambert questions Kimi serving Claude tokens directly to users — natolambert · 2026-10-18
- Don't pick your AI based on influencers: frontier models are nearly equal, choose by UI and multimodal needs — techartist_ · 2026-10-18
- Gemini's Redo with Pro vanished a day after launch, frustrating Ultra users — Artistic-Reviewer406 · 2026-10-18
- TypeSafe raises $870M at $7.5B valuation under a month after Jev model launch — dl_weekly · 2026-10-18
- No benchmark with RSI in its name actually measures RSI capability — mariofilhoml · 2026-10-18