Nanbeige4.3-33B released with open weights and reported benchmark wins
A post says the model’s weights are available under Apache 2.0, with support for a 512K-token context window.
TLDR
A post says Nanbeige released Nanbeige4.3-33B with open weights and strong coding and agent performance. It claims the model beats similarly sized Qwen 3.8 27b and Gemma models on most benchmarks. Another commenter expects those results to be largely true based on past experience with Nanbeige, but cautions that performance could be fragile and prone to “doom looping.”
Combined views
32.9K
7 Sources, first seen ago
