The Batch API is the easiest 50% most teams leave on the table. You submit a file of requests, results come back within a window (up to 24h), and you pay roughly half the per-token price. Nothing about the model changes — same quality, same tokens, half the bill — as long as the work can wait.
—standard / month
—you save / month
—you save / year
Batch cost by model
Same workload, batch price, ranked cheapest first.
| Model | Standard / mo | Batch / mo |
|---|
⚠️ Estimate using reference prices (July 2026) and a list batch discount of ~50%. Exact batch discounts, token rates and the turnaround window vary by provider and change over time — confirm on the provider's pricing page. Batch is per-job async; it does not lower the cost of interactive, real-time calls. ·
Report outdated price →
What batch is — and isn't
The Batch API trades latency for price. You hand the provider a file of requests; it processes them within a window (often up to 24 hours) and returns a file of results, at roughly half the standard per-token rate on both input and output. The model, the prompt and the output are identical — you're only giving up "right now." So it's perfect for work nobody is staring at a spinner for, and useless for anything interactive.
What to move to batch
Good candidates: nightly summarisation, bulk classification or tagging, embeddings backfills, dataset labelling, offline evaluations, and any report generated on a schedule. Keep on the real-time API: live chat, anything a user waits on, and latency-sensitive tool calls. A common pattern is to run the same pipeline both ways — real-time for the interactive path, batch for the overnight reprocessing — and the batch half quietly halves that portion of the bill. Stack it with prompt caching on the repeated context and the savings compound.
Estimating a whole product instead? Use the AI app cost estimator or the prompt caching savings calculator. Building agents? The AI agent cost calculator.
AI Cost Per Resolved TicketAI Cost SavingsLLM Cost OptimizationPrompt Caching SavingsFree Tier Runway
How this calculator works
The Batch API Savings Calculator estimates how much you save by routing asynchronous jobs through a provider's Batch API instead of the standard real-time API. It multiplies your requests per month by the input and output tokens per request to get total token volume, prices that volume at the selected model's rates, and then applies the batch discount — roughly 50% off at OpenAI and Anthropic — to show your standard cost, discounted cost, and monthly savings. The main drivers are request volume, tokens per request, the model's per-token price, and the discount percentage.
The key trade-off is latency for price: batch jobs can take up to 24 hours to return, so this only fits work that can wait — bulk classification, embeddings, evaluations, or offline content generation — not anything user-facing. Before committing, confirm your actual workload tolerates the delay, and check that the token counts you enter reflect real averages, since output tokens are usually the more expensive side and drive most of the total.
Frequently asked questions
How much does the Batch API save?
OpenAI and Anthropic both discount asynchronous batch jobs by roughly 50% on input and output tokens, in exchange for results returning within a window of up to 24 hours instead of in real time. On large bulk workloads that halves the bill.
When should I use the Batch API instead of real-time?
Use batch for any job the user is not waiting on live: nightly summarisation, bulk classification, embeddings backfills, dataset labelling, evaluations. Keep real-time only for interactive chat and anything latency-sensitive.