Poetiq, a startup founded by former Google and DeepMind researchers, has announced that its Meta-System has set a new record on the LiveCodeBench Pro programming benchmark. This API-based intelligent harness enhances coding capabilities of large models without modifying model weights or fine-tuning. Notably, Kimi K2.6's accuracy increased from 50.0% to 79.9%, a 29.9 percentage point gain, while Gemini 3.0 Flash improved by 10 points, outperforming larger models like Claude Opus 4.7 and GPT 5.2 High.
The Meta-System also pushed GPT 5.5 High's score from 89.6% to 93.9%, and helped Gemini 3.1 Pro achieve 90.9%, surpassing Google's Gemini 3 Deep Think at 88.8%. Poetiq emphasizes that their harness allows enterprises to enhance reasoning capabilities without the high costs associated with traditional fine-tuning and deployment of full-scale models.
Poetiq's Meta-System Boosts AI Model Performance by Up to 29.9%
Disclaimer: The content provided on Phemex News is for informational purposes only. We do not guarantee the quality, accuracy, or completeness of the information sourced from third-party articles. The content on this page does not constitute financial or investment advice. We strongly encourage you to conduct you own research and consult with a qualified financial advisor before making any investment decisions.
