Cheapest provider
$10.00 / 1M
OpenAI
Fastest provider (p95)
114 tok/s
OpenAI
Intelligence (composite)
—
Not a chat-completion model
Per-provider performance
Latency / throughput / uptime / price measured across providers over the last 30 minutes of live traffic. This is what proves “sourced cheapest” — Atlas mode draws on these per call to serve the cheapest path that holds quality.
| Provider | Quant | Input $/1M | Output $/1M | Latency p50 / p95 | Throughput p50 / p95 | Uptime 30m | Success |
|---|---|---|---|---|---|---|---|
| OpenAI | full· unknown | $10.0000 | $10.0000 | 7656ms / 30848ms | 100 / 113.6 tok/s | — | 100.0% |
“—” means live telemetry hasn’t accumulated enough recent traffic for that endpoint. “undisclosed” means the provider serves the model but doesn’t expose the quantization label (typically running fp8 / int8 internally).
How Atlas mode sources OpenAI: GPT-5 Image
- Strict mode — pin OpenAI: GPT-5 Image exactly and we pass it straight through, sourced from the cheapest provider above. The same model, no substitutions — currently OpenAI at $10.00/1M.
- Atlas mode — the default. Each call is auto-optimized for the cheapest path that holds quality, at least 5% off going direct from call one and climbing as it ramps. You always see which model served the call and exactly what you saved — thumbs-down anything you don’t like for a full refund.