Z.ai released GLM-5.3 Flash on 26 August 2026, a finished multimodal model under the MIT licence with a published price and a served endpoint, while Baidu released ERNIE-5.5-Flash-Base on 5 November 2026, the end-of-pretraining checkpoint of a model that has not been post-trained, under a bespoke licence with no price and no endpoint. GLM-5.3 Flash is a 320B-parameter model with 18B active per token, 45 layers, a hybrid of linear and sparse attention that Z.ai says cuts attention compute roughly threefold and shrinks the KV cache about 4.4 times relative to the text-only GLM-5.3, a context window of 1,048,576 tokens, output capped somewhere between 128K and 131K depending on the host, text, images and video in with text out, and FP8 weights on Hugging Face running to roughly 328GB. ERNIE-5.5-Flash-Base is 280B total with 14B active, 33.5T pretraining tokens, 43 layers — one dense plus 42 MoE — with 37 layers on compressed HCA attention and six on MLA plus DeepSeek-style sparse attention, hidden size 4,096, 64 heads, 512 routed experts at top-10 plus one shared expert, vocabulary 201,