A frontier Mixture-of-Experts model for deep reasoning, agentic workflows, and tool use, now live on our Canadian fleet alongside GLM-5.2. One OpenAI-compatible API, metered by the token, with your prompts never crossing the border.
DeepSeek-V4-Pro is a large Mixture-of-Experts model built for hard reasoning, long multi-step tasks, and agentic tool use, served entirely from inside Canadian jurisdiction.
A sparse MoE design with 384 experts and 6 active per token, so you get frontier-scale capability with efficient inference.
Deep multi-step reasoning, code, and native tool calling for building agents and copilots on sovereign infrastructure.
Call it as deepseek-v4 on the same endpoint as the rest of the catalogue. Point your existing SDK at ai.spuric.com and switch.
$1 per 1M input tokens and $4 per 1M output tokens. No quote-walls, no minimums, no cross-border data transfer.
DeepSeek-V4-Pro runs alongside GLM-5.2 as our two frontier flagships. Use either by name, or let HARI read each request and route it to the best model automatically, so you never have to manage a model zoo.
Get an API key and call DeepSeek-V4-Pro and GLM-5.2 hosted in Canada, or talk to us about a private deployment. No quote-walls, no border.
