●Singapore endpoint · Data stays outside mainland China · Stripe billing

Your tools, powered by
China's frontier models

Qwen3-Coder and DeepSeek-V4 are live, Kimi is next — one key for all of them. Keep Claude Code, Cline or Cursor exactly as they are: two lines of config swap the engine. Transparent per-token pricing, or a flat monthly plan.

Get early access 2-line setup
OpenAI & Anthropic compatible Low latency across APAC Every request itemized
terminal — Claude Code on Qwen3-Coder
# two lines. that's the whole migration.
export ANTHROPIC_BASE_URL="https://api.gloritoken.ai"
export ANTHROPIC_AUTH_TOKEN="sk-your-key"

$ claude
✔ Connected · model: qwen3-coder-plus
> refactor this module and add unit tests…
4+frontier Chinese models, one key
1Mcontext window · whole-repo aware
~10xcheaper than comparable US models
100%itemized, auditable billing
WORKS WITH YOUR TOOLS

Keep your workflow. Swap the engine.

Anything that speaks the OpenAI or Anthropic API works out of the box — point it at gloritoken and you're running on China's frontier coding models.

Claude CodeNative Anthropic API

Qwen CodeOpenAI protocol

ClineVS Code extension

CursorCustom base URL

Cx

Codex CLICustom provider

RC

Roo CodeOpenAI-compatible

OpenCodeCustom models

CS

Cherry StudioDesktop client

Cb

ChatboxDesktop client

+

Any toolOpenAI-compatible

* Tool names and logos belong to their respective owners; shown for compatibility only, no affiliation or endorsement implied.

MODEL MATRIX

The Chinese coding lineup, first-hand

Supplied through Alibaba Cloud Model Studio (Singapore) and official model-maker APIs — first-hand supply, list-price aligned, no gray-market relays.

QWEN · MODEL STUDIO INTL
qwen3-coder-plusFlagship
Flagship coding model. Best-in-class agentic tool use — the go-to engine for Claude Code.
$1.0 +in /1M
$5.0 +out /1M
1M ctx
QWEN · MODEL STUDIO INTL
qwen3-coder-flashValue
Fast and cheap — completions, high-frequency and lightweight tasks.
$0.3 +in /1M
$1.5 +out /1M
1M ctx
DEEPSEEK · OFFICIAL API
deepseek-v4-proNew
DeepSeek's all-round flagship — strong reasoning and agentic performance.
$0.42in /1M
$0.84out /1M
1M ctx
DEEPSEEK · OFFICIAL API
deepseek-v4-flashValue
The lowest-cost option for completions and high-volume tasks.
$0.14in /1M
$0.28out /1M
1M ctx

* Demo pricing, aligned to upstream list prices (Qwen tiered by context length — "+" marks the entry tier). Kimi and GLM are next in line. We do not relay Claude, GPT or Gemini.

PRICING

A flat plan, or pay as you go

The first subscription that bundles multiple Chinese frontier models. Switch between Qwen and DeepSeek inside Claude Code — one plan covers both.

Starter

$9 /mo
Solo developers · 600 requests / 5h window
  • Unlimited qwen3-coder-flash & deepseek-v4-flash
  • Standard quota on flagship models
  • Works with every coding tool
  • Usage dashboard
Join beta

Pro

$29 /mo
Heavy users · 3,000 requests / 5h window
  • All models, all tiers
  • Model switching inside one session
  • Higher rate limits & concurrency
  • Priority access to new models
Join beta

Credits

List price +
Pay as you go · no minimums
  • Aligned to upstream list prices
  • Stripe checkout, instant top-up
  • Best for CI, backends & batch jobs
  • Volume discounts available
Join beta

* Demo pricing — final plans announced at launch. Subscriptions use a 5-hour rolling window, same convention as major coding plans.

FAQ

Fair questions

Where does my code actually go?

Default routing is our Singapore endpoint, supplied by Alibaba Cloud Model Studio (Singapore) and official model-maker APIs. Your data stays outside mainland China. We don't store your prompt contents and never train on your data — we keep only the usage metadata needed for billing.

How is this different from OpenRouter?

We're specialists, not a catalog. gloritoken only does Chinese frontier models, first-hand — and we're the only place offering a flat monthly plan that spans multiple Chinese model makers. If you loved the GLM coding plan but wish it also had Qwen and DeepSeek in it, that's exactly what this is.

Does it really work with Claude Code?

Yes — natively. We expose a real Anthropic-protocol endpoint (/v1/messages), so Claude Code connects with two environment variables, no proxy hacks. Cline, Roo Code, Cursor and any OpenAI-compatible tool work the same way through /v1.

Swap your engine today

We're onboarding early users now — free trial credits included. API endpoint and billing go live shortly; email to reserve your seat.

liujialiang@gloritoken.com Follow the build on X