Batch LLM Status Tracker

Batch LLM Status Tracker

Use one LLM node for bulk prompting and a second branch for progress-aware side effects. This template demonstrates Heym's Batch API mode on the plain LLM node, including the dedicated batchStatus branch.

What this workflow does

1. Variable node seeds a demo array of prompts
2. LLM node sends that array through Batch API mode
3. STATUS branch maps live progress updates such as pending, processing, and completed
4. Main path reshapes the final per-item results and returns them as the workflow output

Use cases

• Lower-cost bulk prompting via OpenAI's Batch API
• Queue-style AI enrichment jobs
• Progress notifications while the provider batch is still running
• Batch JSON extraction with downstream per-item handling

Setup

Use an OpenAI API credential and a model Heym shows as batch-capable. Replace the demo $vars.promptList array with a dynamic expression such as $input.items.map("item.text") when you turn this into a production workflow.

Notes

Batch mode is LLM-only and text-only. It cannot be combined with image output or image input.