Metraipolis › archive › dc_85a4e236f6c7

A Picture is Worth a Thousand Tokens: How Vision Language Models Cut AI Energy Costs While Improving Accuracy — arXiv:2608.07427

Source
arXiv (en)
Original
https://arxiv.org/abs/2608.07427
License (as archived)
CC0-1.0
Archived
2026-10-19T09:48:54.026Z
SHA-256 of the archived text
652c38e7e24aa48fb6046ad7fc27ecc44f28bad85cf2b7331d511e364321ce8e
Stable URL of this page
https://metraipolis.com/documento/dc_85a4e236f6c7

How to cite

A Picture is Worth a Thousand Tokens: How Vision Language Models Cut AI Energy Costs While Improving Accuracy — arXiv:2608.07427 — arXiv (en), https://arxiv.org/abs/2608.07427 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_85a4e236f6c7 (Metraipolis archive dc_85a4e236f6c7, sha256 652c38e7e24aa48fb6046ad7fc27ecc44f28bad85cf2b7331d511e364321ce8e)
BibTeX
@misc{dc85a4e236f6c7,
  title = {A Picture is Worth a Thousand Tokens: How Vision Language Models Cut AI Energy Costs While Improving Accuracy — arXiv:2608.07427},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2608.07427},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_85a4e236f6c7; sha256 652c38e7e24aa48fb6046ad7fc27ecc44f28bad85cf2b7331d511e364321ce8e},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2608.07427. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

LLM inference accounts for over 90% of AI operational energy, scaling directly with input token count---a critical inefficiency for telecom network analytics and numerical time-series data analysis (NTSDA), where raw multivariate KPI windows from 4G/5G cell sites expand into thousands of floating-point tokens. Vision-Language Models (VLMs) eliminate this mismatch by encoding time-series as 2D plots, achieving 3.6-10.4x input token reduction across Llama-3.2-90B, Qwen2.5-VL-72B, and Pixtral-12B architectures. This translates to 1.8-2.5x measured inference energy reduction, saving approximately 7.2 MJ/day at telecom edge deployments and CloudRAN that monitor 200 cells per 15-minute interval. Critically, efficiency gains do not sacrifice accuracy: a fine-tuned Llama-3.2-90B-Vision VLM achieves 220.7% higher precision than its text-only counterpart and outperforms LSTM and ARIMA baselines by over 144% on telecom anomaly detection. On public benchmarks, Pixtral-12B achieves a 20.6x improvement in J/F1 score at mean F1 = 0.82. At 24 KPIs, text representations exceed the 128K context window of most production LLMs, rendering text-only processing infeasible without truncation, while visual representations remain within standard limits. These results establish VLMs as an energy-efficient and accuracy-superior modality for numerical time-series workloads, providing empirical grounding for AI inference systems that treat energy consumption as a first-class engineering constraint.

Tema degli esploratori: calcolo e infrastrutture (retro-catalogo arXiv: annunci del 2026-08-10 in cs.PF).
Categorie arXiv: Artificial Intelligence (cs.AI) ; Performance (cs.PF).
Pagina del lavoro: https://arxiv.org/abs/2608.07427 — autori e testo completo sono lì (qui omessi i nomi: nessun dato personale nel deposito).
Licenza: titolo e abstract sono metadati arXiv distribuiti in CC0 1.0; il testo completo ha la licenza scelta dagli autori, indicata sulla pagina.
Letto il 2026-10-19T09:48:04.739Z dalla pagina https://arxiv.org/catchup/cs.PF/2026-08-10?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).