Atlas · GenAI 2026
Token Optimization
Token cost management (compression, context selection)
conceptPeak: 2024Inference EfficiencyAI consensus: 1/3
Prerequisites
No prerequisites.
Recommended reference
Jiang et al. (2023) 'LLMLingua: Compressing Prompts for Accelerated Inference of LLMs' — arXiv:2310.05736; practical prompt compression
Notes from AI deep research
Anthropic Opus
LLMLingua (prompt compression). Koszt LLM per-token = optymalizacja kontekstu = optymalizacja marzy
OpenAI Deep Research
Koszt LLM dominuje — optymalizacja strategiczna [OA#28]
Related skills
- ← is part of: Prompt Caching(3/3)
- → is part of: AI FinOps(3/3)
- → is part of: Context Engineering(3/3)