Is NieoToken compatible with the OpenAI SDK?

Yes. NieoToken exposes an OpenAI-compatible REST API. Point any OpenAI SDK or client at the base URL https://nieotoken.com/v1 and use your NieoToken key — no code changes required.

How is NieoToken billed?

Pay-as-you-go. Top up your account and pay only for tokens you actually use. There is no monthly subscription or seat fee.

Does NieoToken support high-volume / batch workloads?

Yes. Traffic is load-balanced across multiple upstream accounts with automatic failover and retry, so sustained high-volume workloads are handled without you managing multiple keys.

Can I send images to vision models?

Yes. Send images as base64 data URLs in the standard OpenAI image_url field to vision-capable models such as DeepSeek V4 Vision.

⚡ Leading China-based LLMs · One key · Pay as you go

Access DeepSeek V4, Qwen & GLM through one API

Q: Which models can I access through NieoToken?

NieoToken currently routes DeepSeek V4 (flash / pro / vision), Alibaba Qwen and Zhipu GLM through one OpenAI-compatible endpoint, with more frontier models available on request. The live model list is always at https://nieotoken.com/docs.

NieoToken gives you a single OpenAI-compatible endpoint for today's best-value frontier models from DeepSeek, Alibaba and Zhipu AI — cheap per-token pricing, high-volume friendly, no subscription lock-in.

Start free in 1 minute → Read integration docs

OpenAI-compatible — works with your existing SDK Pay as you go — no monthly subscription

deepseek-v4-flashqwen3.8-maxglm-5.3deepseek-v4-provision models

nieotoken.com/v1

curl https://nieotoken.com/v1/chat/completions \
  -H "Authorization: Bearer sk-…" \
  -d '{
    "model": "deepseek-v4-flash",
    "messages": [{ "role": "user", "content": "Hello" }]
  }'

→ 200 OK  ·  only pennies per 1M tokens

Why NieoToken

Built for developers who need a lot of tokens without the overhead

💸

Honestly cheap tokens

DeepSeek V4 Flash and Qwen Flash run at a fraction of a cent per thousand tokens. Pay per token — no seats, no minimum spend.

🔁

High volume, no babysitting

Traffic is balanced across multiple upstream accounts with automatic failover and retry. Sustained batches run without you juggling keys.

🔌

Drop-in OpenAI compatibility

Point any OpenAI SDK or app at https://nieotoken.com/v1, swap the key, done. No vendor-specific SDK.

🧠

Leading China-based models, one integration

DeepSeek V4 flash/pro/vision, Alibaba Qwen and Zhipu GLM from a single endpoint — pick per request by model name, plus more frontier models on request.

🖼️

Vision & reasoning included

Send images to vision models via the standard image_url field; switch to reasoning models for hard problems.

📊

Usage you can see

Per-token logs, quota dashboard and token management in your console. Know exactly what each request costs.

Model lineup

Best-value frontier models, one compatible API

Model ID	Developer	Best for
`deepseek-v4-flash`	DeepSeek	Everyday chat, coding, bulk — great value
`deepseek-v4-pro`	DeepSeek	Math, logic, deep analysisreasoning
`deepseek-v4-flash-vision-exp`	DeepSeek	Image understandingvision
`qwen3.8-max · qwen3.8-flash`	Alibaba (Qwen)	Flagship & fast general tasks
`qwen3.x series`	Alibaba (Qwen)	General chat, coding, tool use
`glm-5.3 · glm-5`	Zhipu AI (GLM)	Strong reasoning & long context
`glm-4.x / glm-4.5-air`	Zhipu AI (GLM)	Lightweight, cost-efficient tasks

More models (including OpenAI, Google Gemini and xAI Grok) are available on request — the live, always-updated list is on the integration docs.

Pricing

Pay for tokens. Nothing else.

Value DeepSeek /
Qwen Flash

deepseek-v4-flash & vision
qwen3.8-flash & older
Best for bulk & everyday use
Roughly $0.1–0.7 /M input*

Balanced Qwen /
GLM Flagships

qwen3.8-max, glm-5.3, glm-5
deepseek-v4-pro
Strong reasoning, still great value
Roughly $0.6–3 /M input*

Get started

Flexible Any model
you need

Gemini, Grok, OpenAI on request
One key, switch by model name
Per-token billing, no monthly fee

*Estimates in USD per 1M input tokens — exact per-model rates are on the docs. Top-ups start small; unused balance stays on your account.

FAQ

Questions, answered

Which models can I access through NieoToken?

DeepSeek V4 (flash / pro / vision-exp), Alibaba Qwen and Zhipu GLM are available through one endpoint, and more frontier models can be added on request. The authoritative live list is on the docs page, which updates automatically as models are added.

Is it really OpenAI-compatible?

Yes — the API surface matches OpenAI's chat completions format. Use any OpenAI SDK (Python, Node, etc.), or clients like Chatbox and Cherry Studio, by setting the base URL to https://nieotoken.com/v1.

How does billing work?

Prepaid and pay-as-you-go. Top up a small amount, consume tokens, top up again when needed. There is no monthly subscription, no seat fee, and no forced minimum usage.

Can I run large or batch workloads?

Yes. Requests are load-balanced across multiple upstream accounts with automatic retry and failover, which makes sustained high-volume workloads practical without managing many keys yourself.

How do I send images?

Use vision-capable models (e.g. deepseek-v4-flash-vision-exp) and pass the image as a base64 data: URL in the image_url content part — the standard OpenAI way.

Is my data kept private?

Prompts are forwarded to the selected upstream provider to generate replies (see the upstream's own policy). We don't sell your data. Details in our Privacy Policy.