XerTokenAI
Xerlace Studio presents XerTokenAI

One key for
every AI model.

XerTokenAI is the unified generation gateway. Ship against 120+ models through one OpenAI-compatible API — with smart routing, streaming, and usage that explains itself.

Get your API keyExplore models
  • No credit card
  • 50K free tokens
  • 2-minute setup
120+
Models
99.99%
Uptime
38ms
Routing
2.4B
Tokens / day
xer/autorouted · 38ms
xr_live_••••9f2key verified
1.28M tokenslast 24 hours

live routing mesh

New · GPT-4.1 and Claude Sonnet 4 are live on the gateway

Routing to the models you already trust

OpenAIAnthropicGoogle DeepMindMeta AIMistralDeepSeekxAIAlibaba QwenCohereStability AIOpenAIAnthropicGoogle DeepMindMeta AIMistralDeepSeekxAIAlibaba QwenCohereStability AI

Why XerTokenAI

Infrastructure for generation, not integration work.

Stop maintaining four SDKs and five billing dashboards. XerTokenAI puts every provider behind one contract, then automates the boring parts.

One endpoint. Every provider.

Keep the OpenAI SDK you already use. Change the base URL and switch models with a single string.

request.ts
baseURL: "api.xertokenai.com/v1"
model:   "xer/auto"

→ gpt-4.1        ✓ 38ms
→ claude-sonnet  ✓ 41ms
→ retry          — none

Keys that stay yours

Issue scoped keys per project, cap spend, rotate on demand. Raw provider keys never touch your app.

xr_live_••••••••9f2cactive

$32.40 of $50.00 monthly cap

Smart routing

Requests take the fastest healthy path. Failures reroute before your users notice.

xer/auto38ms
openai71ms
anthropic94ms
fallback120ms

Usage that explains itself

Tokens, latency, spend and error rates per key, model and endpoint — updated in real time, exportable anytime.

1.28M

tokens today

99.98%

success

Streaming, not waiting

Token-by-token SSE with backpressure handling, tool calls and JSON mode preserved across every provider.

Edge-native by default

Terminated in 14 regions, then connected to providers over warm pools. Your p95 stays flat as you grow.

SOC 2 Type II · GDPR ready

From zero to live

Three steps.
Then it is automatic.

Most teams send their first production request in under five minutes. The gateway takes it from there.

  1. 01

    Create your account

    ~40 seconds

    Email and a password. Your first live key is minted the moment your workspace exists — no sales call, no waiting list.

  2. 02

    Point your SDK at us

    3 lines of code

    Keep the OpenAI client. Swap the base URL, paste your XerTokenAI key, and pick any model from the catalog.

  3. 03

    Ship, then watch it scale

    automated

    Streams, retries and fallbacks are handled upstream. Your dashboard shows tokens, spend and latency as they happen.

import OpenAI from "openai";

const client = new OpenAI({
  apiKey: process.env.XERTOKEN_API_KEY,
  baseURL: "https://xerlace.web.id/v1",
});

const stream = await client.chat.completions.create({
  model: "xer/auto",
  messages: [{ role: "user", content: "Automate a launch plan." }],
  stream: true,
});

for await (const chunk of stream) {
  process.stdout.write(chunk.choices[0]?.delta?.content ?? "");
}
Same interface as OpenAI
v1 · stable

Model catalog

Frontier models, one price sheet.

Transparent per-million-token pricing with no markup games. New models land on the gateway the week they ship.

API reference
Filter the catalog
ModelContextInput / 1MOutput / 1Mp95Category
O

GPT-4.1

OpenAI

1M$2.00$8.00412msFlagship
A

Claude Sonnet 4

Anthropic

200K$3.00$15.00468msReasoning
G

Gemini 2.5 Pro

Google

1M$1.25$10.00391msVision
D

DeepSeek V3

DeepSeek

128K$0.27$1.10344msOpen
M

Llama 4 405B

Meta

256K$0.90$2.40302msOpen
M

Mistral Large

Mistral

128K$2.00$6.00288msFast
x

Grok 4

xAI

256K$3.00$15.00501msReasoning
A

Qwen 3 Max

Alibaba

262K$0.60$2.20259msFast

Showing 8 of 120+ models

View full catalog

Unified token accounting

Input, output, cached and reasoning tokens normalised across providers.

Automatic fallbacks

Equivalent-model fallback chains, configurable per key.

No lock-in

Export every request log. Cancel and keep your data.

Built for builders

The API you know.
The reach you don't.

No proprietary protocol, no custom SDK tax. If it works with OpenAI, it works with XerTokenAI.

OpenAI-compatible

The official SDKs work unchanged. Swap baseURL, keep your types.

Typed SDKs

First-party TypeScript and Python clients, generated from the same spec.

Routing controls

Pin providers, set fallback order, and shadow-test a new model safely.

Usage webhooks

Stream billing and budget events into your own systems as they happen.

SSE streaming/Tool calling/JSON mode/Vision
import OpenAI from "openai";

const client = new OpenAI({
  apiKey: process.env.XERTOKEN_API_KEY,
  baseURL: "https://xerlace.web.id/v1",
});

const stream = await client.chat.completions.create({
  model: "xer/auto",
  messages: [{ role: "user", content: "Automate a launch plan." }],
  stream: true,
});

for await (const chunk of stream) {
  process.stdout.write(chunk.choices[0]?.delta?.content ?? "");
}

Base URL

api.xertokenai.com/v1

Auth header

Bearer xr_live_…

Pricing

Start free. Scale when it works.

Every plan includes the full gateway — routing, streaming, analytics and keys. You only move up when your volume does.

Starter

$0forever

For prototypes, side projects and the first production call.

Start free
  • 1 API key
  • 50K tokens / month
  • 12 open + fast models
  • Community support
  • Usage dashboard
Popular

Pro

$29/ month

For teams shipping generative features to real users.

Start 14-day trial
  • Unlimited API keys
  • 5M tokens included
  • All 120+ models
  • Smart routing + fallbacks
  • Streaming & tool calling
  • Email support

Scale

$99/ month

For high-volume products that need control and headroom.

Talk to us
  • Everything in Pro
  • 25M tokens included
  • Custom rate limits
  • Team seats & roles
  • Audit logs
  • 99.99% uptime SLA

Enterprise volume, dedicated capacity or on-prem routing? Talk to Xerlace Studio

Questions

Answers, before you build.

Everything teams ask before their first request. Still curious? We answer in under a day.

hello@xertokenai.com

XerTokenAI is an AI API gateway. You hold one key from us and call any supported model through an OpenAI-compatible endpoint. We handle provider credentials, routing, retries, streaming and metering so you do not have to.

Yes. Point the official OpenAI SDK at our base URL, use your XerTokenAI key, and change the model string. Streaming, tool calling, JSON mode and vision inputs behave the same way.

Plans include a monthly token allowance. Past that, usage is billed per million tokens at the model rate shown in the catalog. You can set hard spend limits per key from the dashboard.

Smart routing detects errors and elevated latency, then retries on an equivalent model from another provider. You can pin an explicit fallback chain per key or opt out entirely.

Prompts and completions are never persisted. We store token counts, latency, status codes and model identifiers for billing and analytics — metadata only.

On Scale and Enterprise you can attach your own OpenAI, Anthropic or Google credentials. Requests then bill directly to your provider while you keep our routing and analytics layer.

Xerlace Studio · Automate Generate

Your next feature
starts with one key.

Create an account, grab a live key, and send your first generation before your coffee is cold.

Create free accountI already have a key

50,000 free tokens · no card · cancel anytime