GLM-5.3-FlashX

GLM-5.3-FlashX is a hosted speed tier of GLM-5.3-Flash, released by Z.ai on 18 September 2026. The vendor's documentation gives it the same 1M-token context window, the same 128K maximum output and the same inputs as the model it is built from — video, image, text and file in, text out — and the only difference it claims is throughput: inference at up to 200 tokens a second for faster responses. Z.ai's own pricing table charges $0.37 per million input tokens, $1.25 per million output and $0.075 per million on cached input, against $0.15, $0.50 and $0.03 for GLM-5.3-Flash on the same page. That is roughly two and a half times the price of the base model for latency alone. No benchmark score is recorded here for exactly that reason: the vendor's claim is that this is the same model, so copying the base model's figures under a second name would manufacture a measurement that nobody made. Buyers should also note what the documentation says about packaging rather than capability — GLM-5.3-Flash is now fully available on the GLM Coding Plan and FlashX is not yet on it, so the faster tier is a pay-as-you-go API cost rather than something a coding subscription absorbs. The 200 tokens a second is a vendor peak, not a measured average, and a workload that is throughput-bound rather than latency-bound gains nothing from the premium.