Help center
FAQ
Find all the answers you care about here
Account
4 questionsTwo easy ways: scan the QR code with WeChat to follow the official account and register automatically; or sign in for the first time with phone number + verification code to register automatically. You'll be placed in the free user group automatically — nothing else needed.
Use phone number + verification code to sign in (no password needed), then change your password under "Console → Settings".
After logging in, go to "Console → API Keys" and click "Create Key" to generate a dedicated API Key. The format is sk-xxx and it can directly replace api_key in the OpenAI SDK.
Three: WeChat scan (recommended — following the account registers you), phone number + password, and phone number + verification code.
Billing
5 questionsPay-as-you-go, billed by Token usage — use whatever you top up. Fees are calculated across four dimensions: input/output/cache-create/cache-read Tokens, with different prices per model. Tokens hit by upstream Prompt Cache are cheaper, significantly reducing cost.
After logging in, go to "Console → Wallet", choose an amount, and pay via Alipay or WeChat Pay. Top-ups arrive instantly with no minimum.
The free route offers a daily number of free calls via the dxnt.com/free entry. Free users get 100 calls/day; higher member tiers get more (Bronze 1000, Silver 2000, Gold 3000). Inviting friends adds extra daily quota, up to +400 calls/day.
Members come in three tiers — Bronze (¥9.9/mo), Silver (¥19.9/mo), Gold (¥29.9/mo) — with higher daily free quotas (1000-3000 calls) and faster rates (higher RPM/TPM).
Plans are fixed call packs, deducted per call within the validity period — suited for users with steady usage. Membership is a monthly subscription offering free quota + rate boosts — suited for long-term heavy use. The two can stack.
API usage
5 questionsClaude (Anthropic), GPT (OpenAI), Gemini (Google), plus domestic models like GLM, Kimi, MiniMax and ERNIE — 70+ models and growing. See the full list on the Pricing page.
The base URL is https://www.dxnt.com/v1, compatible with the standard OpenAI interface format.
Fully compatible. Just replace base_url with https://www.dxnt.com/v1 and api_key with your DX TOKEN key to migrate with no code changes. The Anthropic Messages format is also supported.
Yes. Set stream: true in the request to enable SSE streaming, identical to the OpenAI streaming interface, with automatic failover to a backup channel mid-stream.
Set the environment variables ANTHROPIC_BASE_URL=https://www.dxnt.com and ANTHROPIC_API_KEY=your-key to use it in Claude Code. Note: don't append /v1 to the Base URL; Claude Code appends it automatically.
AIGC creation
3 questionsMainstream text-to-image models like gpt-image-2 and Seedream 4.5. On the "AIGC → Image Generation" page, pick a model, size and quality, enter a prompt, and generate high-quality images.
Yes. Text-to-video models like Seedance 2.0 / 2.5 are available with different resolutions and durations on the "AIGC → Video Generation" page.
Images are billed per image and videos per second, with prices varying by model and resolution. Estimated cost is shown before generation; the free route applies. Failed generations are refunded in full automatically.
Other
3 questionsAll transit is TLS encrypted, API keys are stored with bcrypt hashing, and the system never logs conversation content — only usage data for billing. You can delete keys and usage records at any time.
The platform has a built-in smart scheduler that monitors each channel's health. When an upstream failure is detected, it seamlessly switches to a backup channel within the same request — no user action needed, with recovery time under 1 second.
After logging in, contact online support from the Console page, or follow our WeChat official account for technical support. Members get a priority support channel.