Chat Completions
Supported chat, tool-enabled and real-time workloads.
UVOX FOR OPENAI
Combine safe context reduction with OpenAI-native caching.
Keep your existing SDK, provider account and supported models. Select Auto, Maximum Savings or Zero-Loss Cache and track route, fallback, token and cost results from your dashboard.
Combined modes are beta and account-controlled. Zero-Loss Cache remains the complete-content default until enabled.
OpenAI + UVOX
Base URL https://api.uvox.tech/openai/v1 Authentication Authorization: Bearer YOUR_UVOX_KEY Provider account Connect your OpenAI key securely in the dashboard. Optional request header X-UVOX-Mode: zero_loss_cache
Compatible with common OpenAI workflows
Use familiar application patterns while UVOX applies the allowed engine mode and reports context and cache economics separately.
Responses API
Supported modern OpenAI request structures.
Existing models
Continue using model IDs available in your provider account.
Clear reporting
Review original input, provider input, context avoided, cached input, actual cost, route and fallback status.
Production setup
Use your UVOX key and replace the base URL. Provider credentials can be stored encrypted or supplied securely per request.
Python
from openai import OpenAI
client = OpenAI(
api_key="YOUR_UVOX_KEY",
base_url="https://api.uvox.tech/openai/v1",
default_headers={"X-UVOX-Mode": "zero_loss_cache"},
)Node.js
import OpenAI from "openai";
const client = new OpenAI({
apiKey: "YOUR_UVOX_KEY",
baseURL: "https://api.uvox.tech/openai/v1",
defaultHeaders: {"X-UVOX-Mode": "zero_loss_cache"}
});OpenAI engine modes
The account default can be overridden per eligible request with X-UVOX-Mode.
Auto
Recommended combined beta route. UVOX uses context optimisation only when safety and confidence gates pass.
Maximum Savings
Uses the tenant-configured beta confidence and reduction limits with the same protected-content and original-request fallback chain.
Zero-Loss Cache
Preserves complete request content and coordinates supported provider-native cache affinity.
Request values:
auto, maximum_savings, zero_loss_cache or disabled.PROPRIETARY OPTIMISATION
UVOX keeps the engine private.
Customers receive simple integration, transparent financial measurement and production safeguards without exposure of UVOX’s internal methods.
Simple endpoint integrationEncrypted provider credentialsUsage and savings reportingReal-time workload supportProduction reliability controls
Test OpenAI traffic before switching production.
Start with the free monthly allowance and measure your own workload.
Create free accountIndependently verify OpenAI Zero-Loss Cache
Run direct and UVOX requests with the same exact JSON body. The verifier compares local SHA-256 with UVOX’s prompt-content hash and displays OpenAI provider usage, cached input and distinct response IDs.