Blog
Batch API
Part of the Unbiased glossary: the vocabulary of LLM pricing and infrastructure, defined with receipts instead of adjectives.
A Batch API accepts jobs asynchronously, typically completing within hours instead of seconds, in exchange for a 50% discount on both input and output tokens. Both Anthropic and OpenAI offer it. Batched Claude Fable 5 costs $5/$25 per million tokens, exactly the price of synchronous Opus.
Why it matters
Half off both sides of the meter is the largest flat discount in the market, and it applies to the workloads that least need speed: evaluations, backfills, embeddings pipelines, nightly summaries, dataset labeling. Most teams run these synchronously out of habit and pay double.
The arithmetic nobody runs
Our measured one-shot dashboard task billed $0.1462 on synchronous Fable 5. Batched, the same tokens price at $0.0731, below what we metered on synchronous Opus 5 ($0.1158). If a premium-model workload can wait an hour, you can hold the model and halve the bill. Rates and receipts in the pricing breakdown.
Common questions
How long do batch jobs take?
Both providers target completion within 24 hours and usually finish much faster; overnight windows are typical. The discount is the payment for that flexibility.
Does batch work with caching?
Yes, the discounts stack: batched cache reads bill at the batch discount on the already-reduced cache rate. Heavy pipelines exploit both.
What workloads should never batch?
Anything interactive: chat, agents mid-session, user-facing generation. Batch is for work nobody is waiting on.