Ollama released version 0.14.3 three days ago, introducing several new large language models and text-to-image capabilities. Key additions include Z-Image Turbo, a 6 billion parameter text-to-image model from Alibaba’s Tongyi Lab, designed for generating high-quality photorealistic images. The update also integrates Flux.2 Klein from Black Forest Labs, described as one of their fastest image generation models. For large language models, Ollama 0.14.3 adds GLM-4.7-Flash, which is positioned as a strong 30 billion parameter class model balancing performance and efficiency. Additionally, the release incorporates LFM2.5-1.2B-Thinking, a new hybrid model from the LFM2.5 family.