StealthGPT Platform Updates: December 2023 Edition

Blog, StealthGPT

StealthGPT Platform Updates: December 2023 Edition

December 2023 StealthGPT Updates

Today's blog will go over the various changes to the StealthGPT platform, and boy was it a busy month. We'll start from the top and get to the most interesting updates towards the bottom.

First of all, say hello to an updated StealthGPT landing page, with a robust nav bar (not pictured) and our update our friendly ninja mascot. The StealthGPT landing page and several other pages have been dramatically updated to showcase our growing brand and reputational strength. Expect more aesthetic updates to the landing page, and possibly some changes to the App to come ;)

Next let's talk word count. In the past we were counting both input and output words. So 100,000 words with us would get you roughly 50k output words. We know most users only care about the output words, and did not understand why we counted the input words. We were faced with a dilemma. Cut down the amount of total words to 50,000 on our pricing, which is basically what we have been offering or double the word limit.

Well here at StealthGPT, we cut no corners and we truly value the user experience. So instead of moving our counter to only counting output words and slashing the word allowance per month, we kept the same amount of words. We know that most users will not even realize it but we effectively doubled the word allowance of all users. We stopped counting all input words, and additionally we rolled back the input words into the users allowance. That means in the month of December, any input words you had were given back to you in your monthly word allowance. StealthGPT continues to give the best bang for buck value out there in the space.

This month we created several interesting blogs that covered our competitors, new LLMs, and even StealthGPT itself to see who holds the true title for Undetectable AI. A lot of services will say it but many fail the task. We covered tools like Cognibypass Tool, Phrasley AI, WriteHuman, and new, hot LLM's like Mixtral. Every month we will show off how StealthGPT performs against the best AI Detectors, and we will also be looking into our competitors to see how they hold up as well. We will also test new, popular models to see if they naturally can bypass AI detection. You can count on our team to give you the latest information you need to know about the Undetectable space.

This month we also launched our child project AI Bypass. It is very similar to StealthGPT, we wrote a whole article outlining the differences here: AI Bypass Vs StealthGPT. The short story is AI Bypass is slightly cheaper than StealthGPT, beats all AI Detection, especially Turnitin. However StealthGPT has the most tools and will continue to have more features enabled. You can think of AI Bypass as a more affordable version of StealthGPT that is dedicated solely to bypassing AI Detectors through rephrasing content to be Undetectable. For most people, that's all they need. StealthGPT will become more a tool for power users seeking many usecases for Undetectable AI. StealthGPT has many additional features under development. If you're only using StealthGPT for rephrasing content, AI Bypass might be the perfect fit for you and save you a little bit of cash.

Did you know you can get StealthGPT for a lifetime right now for as little as $50? In case you have not heard we partnered up with AppSumo to offer several different StealthGPT packages that last a lifetime. That's right, for a lifetime. This is perfect for power users or people who can see themselves needed Undetectable AI for many years to come.

You can purchase the App Sumo <> StealthGPT Lifetime Deal here: StealthGPT for Life.

Bonus: The Sumolings love the product with a 4/5 taco score from 11 reviewers and they made the best StealthGPT ad we have ever seen.

Now let's get into the good stuff, two huge updates to the StealthGPT platform:

1. New foundational models and processes: We have built completely new models from scratch over the last month and a half. These models and processes are much better than the previous iteration of StealthGPT outputs. While still not perfect, these models are far better at scoring on: Similarity (how close is the new content to the old content), grammar, spelling, and of course undetectability. StealthGPT's primary goal was to become the most Undetectable AI and we have accomplished that as many articles has showcased. After completing that goal the next mission was to greatly improve our output quality. These new models and processes have brought us significantly closer to our goal but there's a lot of room for improvement.

Here's how the new foundational model scored on the three biggest AI Detectors (out of 20 responses):

  1. Turnitin — 100%
  2. GPTZero -- 87%
  3. Originality -- 80.5%

(100% being undetectable, 0% being very detectable).

The new models are available with various different settings on the Low, Medium, and High mode. The new model is also available through the StealthGPT API.


2. New Scores Interface: Previously all we showed to users was their Undetectability Score (Stealth Score). However we realized we needed to give more data to users to prepare them on how much work their output made need, or if they may want to consider re-bypassing to get a better quality output. Now we have included a similarity, grammar, and spelling score. Keeping these scores increases our transparency with users and also allows us to track output quality to make sure we continuously improve the scores.

Looking Towards January

Well I do not want to tip my cap too much about what is planned in January but expect more aesthetic changes to our landing pages, and the StealthGPT app. Expect output quality to rise, that is our primary mission right now. We are also looking at adding other features to the platform, for example adding multiple language capabilities to our models. Right now they are really good for English, we want to allow other languages once we can assure they are decent quality and undetectable of course. We will continue to test our models and update them if we see AI Detectors catch up to us. We are looking to develop some more ambitious features to the platform, but for now that will remain a secret ;)

I hope you enjoyed reading this monthly update and I look forward to doing these once a month. Thanks for sticking with us.

Written By

Jozef Gherman

Jozef Gherman

Undetectable AI, The Ultimate AI Bypasser & Humanizer

Humanize your AI-written essays, papers, and content with the only AI rephraser that beats Turnitin.