GLM 5.3 FlashX

Command Code's model page for z-ai/glm-5.3-flashx lists the model as a high-speed GLM-5.3 Flash variant with a 1M-token context window, released September 18, 2026, at $0.37 per million input tokens, $1.25 per million output tokens, and $0.07 per million cache-read tokens, yielding an effective agent-loop input rate of The page also provides real-work cost estimates for GLM-5.3-FlashX across representative coding tasks: a quick lookup at roughly $0.002, a 500-line PR review at $0.01, a bug fix in an agent loop at $0.04, a module refactor at $0.08, and a full-repo agent run at $0.18—demonstrating that most of the cost in agent workloa

Superpower Daily reports that Z.AI added GLM-5.3-Flash to its Coding Plan with three times the quota of GLM-5.3, while the faster GLM-5.3-FlashX variant remains API-only and excluded from the subscription plan. Both models are available through Z.AI's API under the model codes glm-5.3-flash and glm-5.3-flashx respectiv Z.AI states Flash has 320 billion total parameters with 18 billion active per task and reports 3.01× less attention computation and 4.44× lower key-value cache usage compared with GLM-5.3 on long inputs. Target use cases span visual coding and browser interaction, document processing, financial research, video understa

Zhipu released GLM-5.3-FlashX on September 18, 2026 as a high-speed serving tier of the GLM-5.3-Flash model, with the vendor officially stating a peak inference speed of up to 200 tokens per second. The tier builds on the previously released GLM-5.3-Flash (which had circulated globally among developers under the name " According to the same coverage, FlashX is positioned to deliver intelligence, price, and speed simultaneously for enterprise developers, with the API going live alongside the announcement. The article frames the speed boost as the result of continued infrastructure investment and inference optimization on Chinese-made

BigGo Finance reports that Zhipu released GLM-5.3-FlashX on September 18, 2026 as the high-speed variant of the GLM-5.3 series, boosting inference speed from the previous 30–50 tokens per second on standard GLM-5.3-Flash to a published maximum of 200 tokens per second—a roughly five-to-six-fold improvement that closes The speed gains are attributed to an inference cluster built on more than 100,000 Chinese-made AI accelerators, with Zhipu noting that hardware efficiency is now comparable to NVIDIA GPUs and that system throughput improved 3.2× within two weeks. InfraAgent, powered by GLM-5.3, participated in infrastructure optimizati

TechFlow reported on September 18, 2026 that Zhipu officially launched GLM-5.3-FlashX, with the API going live under the explicit model identifier "GLM-5.3-FlashX". The piece notes the model had previously been available to global developers under the "Ox Alpha" name and confirms Zhipu's stated peak speed of up to 200 The report attributes both the launch and the infrastructure description directly to Zhipu, distinguishing the creator from OpenRouter or any routing layer. The brief corroborates the model identifier string, the launch date, and the speed claim that other same-day sources also report, providing timely confirmation of

APIMaster's technical breakdown distinguishes GLM-5.3-FlashX from its base sibling: FlashX is the speed-optimized serving tier of the same GLM-5.3-Flash checkpoint (320B total / 18B activated parameters, 1M-token context, 128K max output, native video/image/text/file multimodal inputs, text output), with the API model The article cites Zhipu's list pricing of $0.37 input and $1.25 output per 1M tokens for FlashX against $0.15 and $0.50 for the base Flash tier, and notes that Flash is the open-weights variant included in the GLM Coding Plan (with 3× the GLM-5.3 quota), while FlashX is positioned for throughput-sensitive, low-latency,