A user on Reddit shared their experience with Anthropic's Fable 5.1 model, noting significant differences in token usage based on efficiency instructions. One account, prompted to be token-efficient, retained 45% of its weekly limit after heavy use. In contrast, the other account, without efficiency instructions, depleted its 5-hour limit in under 20 minutes, with Fable 5.1 reporting the creation of 140 agents and the consumption of 11 million tokens. AI
IMPACT Highlights the importance of prompt engineering for controlling AI model resource consumption and cost.
RANK_REASON User experience report on model behavior, not a primary release or research.
AI-generated summary · Google Gemini · from 1 sources. How we write summaries →