Inceptron

The OpenRouter listing for Z.ai's GLM 5.2 (released June 16, 2026, 1M-token context, large-scale reasoning model targeted at long-horizon agent workflows and coding/tool use) lists Inceptron as one of the inference hosts with input pricing of $0.71 per million tokens, output pricing of $2.35 per million tokens, and cac Inceptron's pricing for GLM 5.2 places it in the lower-cost band relative to top-tier providers like Mistral ($1.40/$4.40) and CoreWeave ($0.76/$2.42), but above discount hosts such as DeepInfra's 35%-off rate of $0.4875/$1.56. As with the DeepSeek listing, this evidence is drawn solely from OpenRouter's aggregator tab

According to the OpenRouter marketplace listing for DeepSeek V4 Flash 0731, Inceptron is confirmed as one of multiple inference providers hosting this sparse mixture-of-experts model (13B active / 284B total parameters, 1M context, released July 31, 2026). The Inceptron row shows input pricing of $0.13 per million toke While the listing demonstrates that Inceptron offers DeepSeek V4 Flash 0731 at a mid-range tier (cheaper than DeepSeek's own $0.22/$0.66 listing but more expensive than OpenInference at $0.05/$0.16), the source is a third-party routing aggregator rather than an official Inceptron announcement. No first-party Inceptron