{"slug": "i-shrank-a-428b-model-from-855gb-to-128gb-and-it-still-beats-gpt-5-5-at-coding", "title": "I Shrank a 428B Model From 855GB to 128GB and It Still Beats GPT-5.5 at Coding", "summary": "A 428-billion-parameter open-weight model achieved 59.0% on SWE-Bench Pro, outperforming GPT-5.5's 58.6%, after being compressed from 855GB to 128GB. The model's efficiency and performance mark a significant advance in AI model optimization.", "body_md": "A 428-billion-parameter open-weight model scores 59.0% on SWE-Bench Pro, edging out GPT-5.5’s 58.6% — and the community just finished…\nContinue reading on Towards AI »", "url": "https://wpnews.pro/news/i-shrank-a-428b-model-from-855gb-to-128gb-and-it-still-beats-gpt-5-5-at-coding", "canonical_source": "https://pub.towardsai.net/i-shrank-a-428b-model-from-855gb-to-128gb-and-it-still-beats-gpt-5-5-at-coding-01e653d50fb9?source=rss----98111c9905da---4", "published_at": "2026-06-26 04:02:53+00:00", "updated_at": "2026-06-26 04:11:38.345651+00:00", "lang": "en", "topics": ["large-language-models", "ai-research", "ai-products"], "entities": ["GPT-5.5", "SWE-Bench Pro", "Towards AI"], "alternates": {"html": "https://wpnews.pro/news/i-shrank-a-428b-model-from-855gb-to-128gb-and-it-still-beats-gpt-5-5-at-coding", "markdown": "https://wpnews.pro/news/i-shrank-a-428b-model-from-855gb-to-128gb-and-it-still-beats-gpt-5-5-at-coding.md", "text": "https://wpnews.pro/news/i-shrank-a-428b-model-from-855gb-to-128gb-and-it-still-beats-gpt-5-5-at-coding.txt", "jsonld": "https://wpnews.pro/news/i-shrank-a-428b-model-from-855gb-to-128gb-and-it-still-beats-gpt-5-5-at-coding.jsonld"}}