Merge Gateway Launches Batch Inference at ~50% of Standard Prices

shensi · x · 2026-09-25

Merge Gateway launched batch inference: submit up to 10,000 requests in one job, routed to provider batch APIs (Anthropic, OpenAI), finishing within 24 hours at typically 50% of standard pricing. Suited for evals, backfills, classification, document processing, and embeddings. Billing is per-line with individual retries — errored, expired, or cancelled lines cost nothing. Docs expose an /llms.txt index for agents.

Original post →

More from Infra

Infra channel →