Chinese LLM API Relays
59 relays listed
Chinese large language models — DeepSeek, Qwen, Kimi, GLM, and others — are inexpensive and improving fast, and a lot of relays put them behind the same API surface as Western models, so switching between a Chinese model and GPT or Claude is just a model-name change in your request.
Every relay below carries at least one Chinese model. Since these models are already cheap at the source, unified API access and reliability matter more here than squeezing out the lowest multiplier.
Relays59
Verify each provider yourself, and test with a small deposit before topping up more
NovaAPI is a global AI token relay platform offering unified access to OpenAI, Claude, Gemini, and leading Chinese LLMs through low-latency, always-on routing. Pricing is fully transparent, with API calls discounted up to 15% off list rates, and referring a friend earns ongoing rewards.
OpenAI
Claude
Gemini+1
Runs a full, unthrottled GLM 5.2 deployment on a dedicated H200 cluster — no hidden rate limits, built for agent workloads, batch jobs, and long-document development. One-stop access to the latest Chinese models including Kimi K3, Seedance 2.0, and Vidu Q3 Pro, with free credits for new users, a dedicated student discount, and open reseller partnerships.
A full-modality AI API supply layer aggregating 500+ models — Sora 2, Veo 3.1, Kling 3.0, Seedream 5.0, GPT-5, Claude, and more. Supports prepaid top-ups, dedicated enterprise channels, and revenue-share partnerships, serving relay platforms, AI app developers, and teams building for global markets.
GrokGemini+2
OpenAIClaudeGemini+1
OpenAIClaudeGemini+2
OpenAIClaudeGemini+2
OpenAIClaudeGemini+2
OpenAIClaudeGemini+2
OpenAIClaudeGemini+2
OpenAIClaudeGemini+2
OpenAIClaudeGemini+2
OpenAIClaudeGemini+2
OpenAIClaudeGemini+2
OpenAIClaudeGemini+2
GrokGemini+2
OpenAIClaudeGemini+2
OpenAIClaudeGemini+2
OpenAIClaudeGemini+2
OpenAIGeminiClaude+2
OpenAIClaudeGemini+2
OpenAIClaudeChinese LLMs+1
OpenAIClaudeGemini+1
OpenAIClaude+2
OpenAIClaudeGemini+1
OpenAIClaudeGemini+2
OpenAIClaudeGemini+1
OpenAIClaudeGemini+2
OpenAIClaudeGemini+2
OpenAIGeminiGrok+2
OpenAIClaudeGemini+2How to choose one
Pick a relay that puts Chinese and Western models behind the same API — swapping models becomes a one-line change in your code.
Confirm the relay actually carries the specific model you need (e.g. DeepSeek-V3/R1) rather than assuming "Chinese models" is a single bucket.
Since per-token pricing is already low, reliability and concurrency limits matter more than shaving another fraction off the multiplier.
Frequently asked questions
Do I need a relay to use Chinese LLMs?
If you're mixing Chinese and Western models, a relay that unifies both behind one API is the most convenient path; for Chinese-only usage you can also go directly to the official API.
Do relays support DeepSeek?
Most do — check each station card for its exact model lineup, since coverage varies by relay.
Are Chinese LLM relays cheap?
The underlying models are already inexpensive, so a low-multiplier relay is a good deal for high-volume use.
Browse relays by category
Find the right AI API relay by model, payment method, or sign-up bonus.