GLM-5.3-Flash: the same million-token window and tool calling as the flagship, at a twentieth of the price.
Added
- GLM-5.3-Flash on both international Z.AI providers. Same 1M context, same
tool calling, and it takes the same graded thinking control — Flash also refuses a disabled thinking block, so it goes through the same machinery as 5.3 rather than the on/off path. Listed at $0.15 in / $0.50 out per 1M; the 50% launch promotion is deliberately not what the cost tracker records, since a promotional rate understates every session the day it ends. China gateways keep 5.2 for the same reason 5.3 is not offered there.
Fixed
kimi-k2.7-code-highspeednow has a rate. It was exempted because Moonshot
published no distinct high-speed price; they now do — double the standard model across every token category. The exemption is removed rather than left as a dead suppression that would hide the next gap.