[AINews] Kimi 3.1 Pro 104BA-2.8T: Fable-5.1-Class Open Weights narrows Closed-AI lead to 1 month... and Pro Highspeed API with 2-3× speed
Kimi K3.1 Pro becomes first open-weights model to beat Claude Fable 5.1
“Kimi K3.1 Pro is a post-trained version of Kimi K3, with substantially larger post-training compute. To enable large-scale joint RL across hundreds of downstream domains, we optimized every layer of the stack — the RL framework, long-context inference, training efficiency, sandbox infrastructure, rollout scheduling, and a shared platform for environment quality checks — so that each increment of RL compute delivers stable, predictable returns.”
Moonshot AI's Kimi K3.1 Pro, a 2.8T parameter open-weights MoE model, has surpassed Claude Fable 5.1 on benchmarks, marking the first time an open-weights model has beaten a top closed-source frontier model. The model achieves this through massively scaled post-training compute with joint RL across hundreds of domains, rather than architectural changes. This narrows the closed-to-open performance gap to roughly one month, a significant milestone that validates earlier predictions about open-weights models catching up to frontier closed models by year-end 2026.