Skip to content
  • 🧰 Features
    • ✨ Features
    • πŸ—ƒοΈ Collections
    • πŸ“„ Prompts
    • πŸ“ Templates
    • 🧠 Analyzers
  • πŸ—ΊοΈ Roadmap
    • πŸ›£οΈ Roadmap
    • πŸ“£ Product Updates
  • πŸ’³ Pricing
    • πŸ’² Pricing
    • ↩️ Refund Policy
  • βš–οΈ Legal
    • πŸ“œ Terms of Service
    • πŸͺ Cookie Policy
    • πŸ” Data Processing Agreement
  • 🏠 About
    • πŸ‘₯ About Us
    • βœ‰οΈ Contact Us
    • ❓ FAQ
Signup
Login
Product Updates

No Credits Charged for Empty AI Generations

08/20/2026

Version: 2.5.1, Date: August 20, 2026

Empty AI generation credits should only ever be spent on something you actually receive. This release fixes a case where a generation could report success, charge you a credit and return nothing at all, and it raises the output limits that were quietly truncating long results.

Prompt Studio

  • Improved: When a generation genuinely produces nothing, both card generation paths now return a clear error instead of a false success message, and the message ends by telling you that no credits were used.
  • Fixed: Custom card generation could report “Generated 1 prompt!” while returning no prompt at all. The empty result flowed through as if it were valid, so the success message appeared after the credit had already been deducted. Card generation now checks that it actually received something before returning a result.

Across the App

  • Improved: Reasoning models are now given extra headroom on top of the budget the app requests. The underlying cause of the empty results was that these models count their own internal reasoning against the same limit as the text you see, so a request could be spent entirely on thinking with nothing left for the answer.
  • Improved: If you use your own API key, an empty generation is now reported as exactly that. It used to be mistaken for a key failure, which offered you a pointless retry using platform credits.
  • Improved: Output limits for AI generation were duplicated across five different engines at different values, some as low as 2,000 tokens, which meant a large multi-output card could be cut off mid sentence. They are now managed in one place and raised to 8,000 across the board. Because billing is based on what is actually produced, and models stop when they are finished, the higher ceiling costs you nothing.
  • Fixed: No credits are charged for an empty generation, anywhere in the app. The check now sits above the point where usage is recorded, across every model provider, so an empty result can never reach your balance.
  • Bug Fixes
  • Improvements
  • Infrastructure

Post navigation

Previous
Next

Search

Categories

  • Articles (4)
  • Product Updates (143)

Recent posts

  • Blueprint Composer Gets Locked Default Sections
  • Topic Pack Focus: Choose the Stages Your Pack Covers
  • Topic Popup Now Names the Topic You Are Editing

Tags

Bug Fixes Deprecated Removed Improvements Infrastructure New Features Platform

Related

Product Updates

Blueprint Composer Gets Locked Default Sections

09/13/2026

Version: 2.33.0, Date: September 13, 2026 Highlights The blueprint composer had two pieces that could be moved and edited but should never have been: Main Idea and AI Model sat in Blueprint Sections alongside the sections you actually assemble, and the preview offered a text box for a main idea that the person running the […]

Product Updates

Topic Pack Focus: Choose the Stages Your Pack Covers

09/13/2026

Version: 2.32.0, Date: September 13, 2026 Highlights Topic pack focus is a new, optional section in “Add topics” that tells generation what the pack is for. Leave it closed and everything works exactly as it did. Open it and you choose the stages your topics should cover, which turns a general list into one built […]

Product Updates

Topic Popup Now Names the Topic You Are Editing

09/12/2026

Version: 2.31.4, Date: September 12, 2026 The topic popup did not tell you which topic you had opened, which is easy to lose track of in a large pack. It does now, and the A to Z sorting is split so you can order topics and subtopics independently. Building Blocks

Product Updates

Subtopic Generation Stays on Topic and on Count

09/12/2026

Version: 2.31.3, Date: September 12, 2026 Subtopic generation inside “Edit topic” now does what the name suggests. Generating from that window used to be able to create new topics rather than adding to the one you opened, and the count you set was a starting point rather than a result, because anything filtered out as […]

Our advanced prompt engineering opens pathways to AI mastery, where systematic thinking transforms creative potential.

Features
  • Library Explorer
  • Prompt Explorer
  • Prompt Builders
  • Prompt Analyzers
General Info
  • Subscription Pricing
  • Refund Policy
  • Roadmap
  • Product Updates
  • Prompt Engineering Articles
Get in touch
  • Contact Us
  • Frequently Asked Questions
  • About Us
  • Terms of Service
  • Privacy Policy

Β© 2026 EnginifyAIβ„’