About us

About DX TOKEN

Making AI model calls simpler, more reliable and more affordable

Our mission

DX TOKEN is dedicated to building the smartest AI model relay gateway. We believe developers should focus on building great products rather than wrestling with API compatibility, channel stability and cost optimization. A unified interface, smart routing and seamless switching — let AI calls be simple again.

Five-Layer Defense Architecture

From routing to streaming, full protection for the reliability of every AI call

Layer 1

Smart Routing

Dynamically selects the optimal channel based on realtime load and latency, ensuring each request takes the fastest path

Layer 2

EMA Channel Scoring

Exponential moving average algorithm evaluates channel health in real time, auto-degrading unhealthy channels to boost availability

Layer 3

Upstream Cache Passthrough

Automatically passes through Prompt Cache from OpenAI/Anthropic/Gemini upstreams, reducing latency and cost

Layer 4

Per-Request Failover

Auto-retries backup channels within a single request, invisible to users, with recovery time under 1 second

Layer 5

Mid-Stream Switching

Detects upstream anomalies during streaming and seamlessly switches to a backup channel to continue output with zero interruption

Core values

Extreme performance

Millisecond routing decisions and smart load balancing ensure each request takes the optimal path

Secure & reliable

End-to-end encryption, multi-layer disaster recovery, no data at rest — protecting every call

Developer first

OpenAI-compatible format, zero-code migration, complete docs and SDKs

Global coverage

Multi-region deployment with nearby access for low-latency availability worldwide

FAQ

A Token is the basic unit AI models use to process text. Roughly 1 Chinese character ≈ 1-2 Tokens, and 1 English word ≈ 1-1.5 Tokens. Both input and output are billed by Token count.
Pay-as-you-go, billed on actual input/output/cache-create/cache-read Tokens across four dimensions. Prices vary by model; see the pricing page. Tokens hit by upstream Prompt Cache are cheaper, significantly reducing cost.
OpenAI (GPT-4o, o1, etc.), Anthropic (Claude Sonnet 4, Claude 3.5, etc.), Google (Gemini 2.0 Flash, Gemini 1.5 Pro, etc.) and the Mistral family, with more being added.
All transit is TLS 1.3 encrypted, API keys are bcrypt-hashed, and request logs never record sensitive content. The system is ISO 27001 certified and supports private deployment.
DX TOKEN has five layers of defense: smart routing picks the optimal channel, the EMA algorithm evaluates channel health in real time, per-request failover switches to a backup channel within 1 second, and mid-stream switching keeps the experience uninterrupted.
Yes! Sign up to get 1M Tokens of free quota, supporting base models like GPT-3.5 and Claude Haiku. No credit card required to start.
Fully compatible. Just replace base_url with https://www.dxnt.com/v1 to migrate with no code changes. It also supports the Anthropic Messages format.
Free users get community support, Pro users get priority support, and Enterprise users get dedicated 24/7 support. You can also email fjzi@qq.com.

Contact us

Email

fjzi@qq.com

Follow our official account for the latest updates and technical support

Join us

Build the future of AI infrastructure with us

Get started