Batchrouter
Start free
Compatibility

OpenAI Batch JSONL, unchanged

Batchrouter accepts the same JSONL format as the OpenAI Batch API. Existing files work as-is.

Get an API keyQuickstart
Input

JSONL

drop in existing files

Endpoints

3

chat · embeddings · responses

Output

JSONL

preserves custom_id

Input format

batch.jsonl

Each line is a JSON object with custom_id, method, url, and body. Supported URLs include /v1/chat/completions, /v1/embeddings, and /v1/responses.

input.jsonl
{"custom_id":"r-1","method":"POST","url":"/v1/chat/completions","body":{"model":"gpt-4o-mini","messages":[{"role":"user","content":"Hi"}]}}

Output format

output.jsonl

Results return as JSONL with one line per input, preserving your custom_id. Each line contains either a response body or a structured error, matching the OpenAI Batch shape.

result.jsonl
{"id":"resp_01","custom_id":"r-1","response":{"status_code":200,"body":{"choices":[{"message":{"role":"assistant","content":"Hello"}}]}},"error":null}

What changes

  • Base URL becomes https://api.batchrouter.dev.
  • Authorization uses your Batchrouter API key.
  • Submission is two steps: quote, then accept — locks the price before any provider work starts.
  • Model names are routed across eligible providers — equivalents are quoted automatically.

Already on OpenAI Batch?

Point your existing JSONL at Batchrouter and ship the same file.

← Back to docsQuickstart