Test any model. See real-time cost and guardrails.
Interactive in-browser workbench for Live models across Anthropic, Google Vertex, and OpenAI. Stream live tokens, verify PII redaction, test failovers, and export SDK code.
Your key is never sent to nRouter servers for storage. Master keys are isolated and disallowed on inference paths.
Evaluate customer order history, determine warranty coverage, and draft a concise approval.
Customer verified under Enterprise SLA (Account #9941). Warranty valid through Dec 2026. Replacing failed power distribution unit under warranty protocol with priority 24h courier dispatch.
import OpenAI from 'openai';
const client = new OpenAI({
baseURL: 'https://api.nrouter.ai/v1',
apiKey: process.env.NROUTER_API_KEY,
});
const stream = await client.chat.completions.create({
model: 'claude-sonnet-4-5',
messages: [{ role: 'user', content: 'Evaluate customer order history, determine warranty coverage, and draft a concise approval.' }],
stream: true,
});
for await (const chunk of stream) {
process.stdout.write(chunk.choices[0]?.delta?.content || '');
}Zero Server Key Storage
Your virtual key is held in browser sessionStorage only. Tab close cleans the memory footprint entirely.
Streaming SSE Live
High-frequency Server-Sent Events with live token counts, TTFT timers, and cancellation support.
Micro-Dollar Cost Telemetry
Every model response returns exact token breakdowns and calculated spend at verified provider rates.
14 Code Generators
Export copy-paste integration snippets in Python, Node.js, Go, PHP, Java, Ruby, and cURL instantly.
Ready to route production workloads?
Add a card, load $5, and start with $15 in credits. Use the OpenAI SDK you already know with 2 lines of configuration.