Batch Runs

Process dozens or hundreds of inputs through any flow at once — manually or via CSV import.

AI Automation Batch Processing Flow Editor

When you need to run a flow dozens or hundreds of times, doing it one by one through the chat window is not practical. Batch Runs solve this by letting you define all your inputs upfront and process them through the flow simultaneously. Whether you are generating product descriptions for an entire catalog, enriching a list of leads, or updating content across multiple URLs, Batch Runs turn a repetitive manual task into a single click.

Tip: Forget the chat output. Couple batch runs with integrations to fully automate your workflows. Output directly into WordPress, HubSpot or anywhere else so that you don’t have to copy the outputs manually.

Popular Use Cases

Batch Runs work with any flow. Here are a few examples of what you can do:

Content generation — Feed a list of product names and specs into a product description generator and get polished copy for every item at once.
Content updates — Pass a list of blog post URLs to a content improver flow and get refreshed versions of each article.
Lead enrichment — Import a CSV of company names or contact details and let a research flow look up additional information for each entry.
Data extraction — Run a scraping or summarization flow against a list of URLs to pull structured data from each page.
Translation — Send a batch of text snippets through a translation flow to localize content in bulk.
SEO analysis — Analyze a list of pages for keyword coverage, meta tag quality, or content gaps in a single run.

How to Use Batch Runs

Step 1: Open the Batch Interface

Navigate to the flow you want to run in batch. In the top navigation bar of the flow editor , click Batch. This opens the Batch Runs panel.

Opening Batch Runs from the flow editor top bar

Step 2: Define Your Inputs

Batch Runs work by processing rows of inputs. Each row represents one execution of the flow. The inputs depend on what the flow expects as input.

For example, if you are using a product description generator, each row might contain a product name and key specifications like “iPhone 17 Pro Max 256GB Black”. If you are using a blog content improver, each row would contain the URL of an article you want to update.

You can add rows manually by typing directly into the input table, or you can import them in bulk as .csv.

Step 3: Import a CSV (Optional)

If you already have your inputs in a spreadsheet, you do not need to type them one by one. Click the Import CSV button and select your file. FlowHunt will populate the input rows automatically, mapping the CSV columns to the flow’s input fields. This is especially useful when working with large datasets.

Step 4: Set Concurrency

Before running, set the concurrency — the number of inputs that will be processed at the same time. The default is 3, which means three rows will run in parallel. You can increase this for faster throughput or decrease it if the flow relies on rate-limited external APIs.

Step 5: Run the Batch

Once your inputs are defined and concurrency is set, click Run Pending. You will see the rows start processing based on the concurrency level you configured. Rows that are waiting will show as pending, active rows will show as running, and completed rows will be marked as done.

Step 6: Monitor Progress and Retrieve Outputs

While the batch is running, you can check on the progress of individual rows at any time by hovering over the row and clicking “View session history”:

Once a run is completed, you’ll get to see the output artifacts for each row. These are the documents and data the flow produced for that input.

For example, in a product description generator flow, each completed row might produce several artifacts: the final product description, a product research summary, a competition comparison, and an SEO analysis. Simpler flows will output just the final result.

If a run fails or you’re not happy with the output, investigate it via the “View session history”. You can duplicate and adjust the row or retry it. Both of these options will be revealed on hover.

Accessing Past Batches

All your batch runs for a flow are saved in the batch history. They are not removed when you close the editor or navigate away. You can always come back to a flow and review the results of any previous batch.

Step 7: Start a New Batch

When you are ready to process a new set of inputs, click + New Batch and repeat the process. Your previous batches remain accessible in the history alongside the new one.

Frequently asked questions

Batch Runs let you process multiple inputs through a single flow at once, instead of running each one individually through the chat. You define your inputs as rows, either manually or by importing a CSV, and FlowHunt runs several sessions in parallel.

Yes. Instead of typing each input manually, you can import a CSV file and FlowHunt will populate the input rows automatically.

Concurrency controls how many inputs are processed at the same time. The default is 3, meaning three rows run in parallel. You can increase or decrease this depending on your needs and available resources.

Yes. All batch runs for a flow are saved in the batch history. They persist even after you close the editor, so you can always come back to review results or retrieve outputs.