Hong Kong low latency
Edge nodes in Hong Kong give overseas developers stable, fast access to mainland China models — no VPN, no throttling.
Gantriq routes through Hong Kong for low-latency, stable access to DeepSeek, GLM, Kimi, Qwen, Seedance video and more. OpenAI-compatible — swap your base_url and ship.
Hardcore reliability, zero migration friction, transparent pricing.
Edge nodes in Hong Kong give overseas developers stable, fast access to mainland China models — no VPN, no throttling.
Keep your SDK and code. Point base_url at Gantriq and every chat, vision and embeddings call just works.
Pay-as-you-go in US dollars. Per-token and per-request pricing shown up front. No minimums, no surprises.
Streaming, function calling, video generation and a clean console. Built by people who live in the terminal.
Text-to-video and image-to-video models from ByteDance and Alibaba. Generate cinematic footage at scale.
Reasoning, chat, long-context, multimodal — OpenAI-compatible API for every Chinese frontier model.
New models added as they launch. Same API, same billing.
View full catalog →Three steps. No SDK rewrite.
Sign up and grab an API key from the console. Free starter credit included.
Point your existing OpenAI client at https://api.gantriq.com/v1
Set the model name and send. Streaming, tools and vision all supported.
from openai import OpenAI client = OpenAI( base_url="https://api.gantriq.com/v1", api_key="sk-gq-...", ) resp = client.chat.completions.create( model="deepseek-chat", messages=[{"role": "user", "content": "ping"}], ) print(resp.choices[0].message.content)
Free credit to start. Scale on transparent USD pay-as-you-go pricing.