One API. Three ways to send work.
Send an immediate chat request, queue one at batch price, or submit a bulk workload. Compare published lanes and set requirements before work starts.
Choose how your work arrives
Direct chat completions
The standard OpenAI request body, answered right away. Set "stream": true to get the reply as server-sent events.
The provider's direct price plus the service fee. The maximum cost is held in your wallet first and settled on actual usage.
One request, queued as a batch
The same request, run as a one-item batch. You get 202 and a job id right away, then read the result from GET /v1/chat/completions/{id} or a webhook.
The batch price plus the service fee, with no separate quote call.
Bulk batches
Many requests at once, from the native API or a JSONL file. An existing OpenAI Batch client works with a base-URL swap.
The batch price plus the service fee, estimated before dispatch.
Quote, route and receive
Prepare the workload
Send items through the native API or use OpenAI Batch-compatible JSONL. Keep a stable customer item identifier for each result.
Compare eligible lanes
Choose your model and privacy requirements. Quotes expose estimated charges and lane rejection reasons before dispatch.
Review the results
Track batch status, then retrieve results or artifact URLs. Handle item errors alongside successful results.
Receive completion signals
Poll the batch endpoint or provide a webhook URL. Verify the delivery signature before trusting a notification.
Read the data declarations
Compare declared privacy tiers, processing countries, retention and training settings. A broad region label alone is not a residency guarantee.
Settle on actual usage
Estimates reserve credits before dispatch. Receipts replace estimates with actual usage charges, and unused reserved credits return to your wallet.
OpenAI-Batch compatible
Point your existing OpenAI batch client at Batchrouter and keep your JSONL format.
Transparent pricing
Every quote itemizes provider cost, Batchrouter fee, and discount. Unselected lanes return rejection reasons.
SLA tiers
The API reports which tiers this environment offers. Published windows are: Priority (4h), Standard (24h), Relaxed (48h), Flex (72h) or Economy (7 days), or Realtime (best effort) for a direct chat request. Batchrouter routes to lanes that meet the deadline, and slower tiers cost less.
Evaluation sweeps are coming soon
The planned evaluation workflow runs your sample across candidates and grades the results against your quality bar. Current public rankings report provider observations and published catalogue information.