Build the AISoftwareFactory Together

Access frontier AI models, open-source inference, image generation, and video models through Relay.

Run AI in One Call Open Native.Builder

FRONTIER ACCESS

WHY RELAYGPU

Built for teams who ship at scale.

Drop-in compatible

Change base_url. Nothing else. Every OpenAI-compatible SDK works out of the box.

99.9% target uptime

Multi-provider routing. One provider goes down, requests reroute automatically. Your code never changes.

All systems operational

No subscriptions

Pay per token. No monthly minimums, no seat fees. Top up once and every token bills at our discounted rate.

One key, frontier models

Claude, GPT-5, DeepSeek, Llama, Qwen, plus image, video, and audio. One API key, zero juggling.

claude-sonnet-4-6gpt-5deepseek-r1llama-3.3-70bflux-2-pro+35 more

ISO 27001 · SOC 2

Enterprise security from day one. EU Sovereign tier adds GDPR, zero retention, and no CLOUD Act exposure.

GDPRSOC 2ISO 27001

Live pricing API

Every price at /v2/pricing. Build cost dashboards, alerts, and budget controls into your infra.

View live prices

INTEGRATION

One line.Every framework.

Same workflow.
One RelayGPU endpoint.
Multiple models behind it.

See all model prices ↗

import axios from 'axios'
const { data } = await axios.post('https://relaygpu.com/v2/ollama/api/chat', {
    model: 'gpt-oss:20b',
    messages: [{
        role: 'user',
        content: 'Break down the pros and cons of decentralized GPU compute.'
      }],
    stream: false,
    think: 'low'
  },
  { headers: { 'X-API-Key': process.env.RELAY_API_KEY } }
)
console.info(data.message.content)