Question 1

What is PromptBatch?

Accepted Answer

PromptBatch is a batch processing platform that allows you to send multiple AI prompts at once to various AI platforms (GPT, Claude, Gemini) and receive results via email. It's designed for users who need to process large volumes of prompts efficiently.

Question 2

How does batch processing work?

Accepted Answer

Upload a CSV file with your prompts, select your preferred AI platform and model, configure output tokens, and submit. For batches with 50+ prompts, processing happens in the background and you'll receive an email notification when results are ready. Smaller batches are processed immediately.

Question 3

What AI platforms are supported?

Accepted Answer

PromptBatch currently supports OpenAI GPT models, Anthropic Claude models, and Google Gemini models. You can select your preferred platform and specific model when processing prompts.

Question 4

What is the Free Plan?

Accepted Answer

New users receive a 2-week free trial with 500,000 tokens. The plan expires when either the time limit or token limit is reached, whichever comes first. After that, you can upgrade to Starter ($25/month) or Professional ($100/month) plans.

Question 5

What are tokens and how are they counted?

Accepted Answer

Tokens are the units AI models use to process text. Roughly, 1 token equals about 4 characters or 0.75 words. Both input (your prompts) and output (AI responses) consume tokens from your balance.

Question 6

How long does batch processing take?

Accepted Answer

Batch processing typically completes within 0-24 hours depending on the size and complexity of your requests. You'll receive an email notification as soon as your results are ready for download.

Question 7

What CSV format should I use?

Accepted Answer

Your CSV file should have a column containing your prompts. The system will automatically detect the prompt column. You can include additional columns for context or reference that will be preserved in your results.

Question 8

Can I use my own API keys?

Accepted Answer

No, you cannot add your own GPT or Claude API keys.

Question 9

How do I download my results?

Accepted Answer

Once processing is complete, you can download results from the Jobs History page. Results are available as CSV files and can be downloaded individually or as a ZIP file for multiple batches.

Question 10

What happens if a prompt fails?

Accepted Answer

Failed prompts are logged with error details. You can view the success and failure counts for each batch in Jobs History and retry failed prompts if needed.

Question 11

When is entering a base prompt helpful?

Accepted Answer

While a base prompt is optional, it is helpful when you want to give the LLM more context on top of the exact prompt you entered in your spreadsheet. eg., You exact prompt might say "write a 200 word description for X" while a helpful base prompt might say "you are an archeologist teaching post-graduate archeology students".

Note, when you add a base prompt, it will add to your total token count.

Question 12

When should I select adding reasoning?

Accepted Answer

Reasoning improves answers only when the task requires decisions; for simple creative generation, it’s the prompt—not the reasoning—that determines quality.

This is why many high-volume content teams deliberately disable reasoning for production runs—it preserves creativity and speed.

Reasoning adds value if the task becomes multi-constraint or decision-heavy.

Frequently asked questions

Still have questions?

Support for Leading AI Models