Qwen 2.5 72B Aces ACT Exam with Perfect Reading Score
on_line187 · reddit · 2026-08-20
The author ran the Qwen 2.5 72B (Q8 GGUF) model on dual RTX 3090s to take official ACT practice tests using its vision capabilities to read PDFs directly. Out of 342 questions across two tests, the model answered 326 correctly (95.3%), achieving a perfect score in Reading and composite scores of 36 and 34 (max 36), rivaling top human performance. The model took about 88 minutes per test.
More from Models
- Zhipu GLM-5.3 DeepSWE Score Jumps to 69 via Post-Training Scaling — haider1 · 2026-08-20
- Cohere study finds 14B models outperform larger ones in linguistic reasoning challenge — Cohere_Labs · 2026-08-20
- Train a 135M parameter LM in 2.5 hours: A full journey — EAccelerate_42 · 2026-08-20
- Ornith-1.5-9B trends on Hugging Face with image-text-to-text support — ornith-ai · 2026-08-20
- Refusal-removed Qwen2.5-72B runs on Mac, zero refusals with high risks — petrusenko_max · 2026-08-20
- Tencent Prepares Larger Multimodal Model Hy4 for Release — zephyr_z9 · 2026-08-20