CheapGPT
more convenient and cost-effective way to use the GPT model.
ChatGPT is very powerful but log in and authenticate with an OTP every time I commute has become exhausting. Every time I want to find something, I have to open a new tab, go to the site, locate a chat in the list, enter it, and then type my message.
This has made me really tired. Considering it costs $22 a month, my usage has significantly decreased.
looking for alternatives for people who want to use ChatGPT more cost-effectively. By utilizing the OpenAI API, you can access GPT models, allowing you to use them conveniently at a relatively lower cost, with high efficiency and accessibility. Use it easily, as much as you need, whenever you want.
Benefits
- Just open and use - You DON'T NEED Many steps for something to do
- Clean desktop - You DON'T NEED Another desktop application
- No more Login and OTP - You DON'T NEED Signin everytime
- Save your cost - You DON'T NEED Pay when you are not use ChatGPT
Usage
- Right click extension icon and select "options" menu
- Input your Open AI API API Key and Save options
- Select Model Optional. Default 4o-mini
- Open popup to click extension icon or
Ctrl + Shift + X
- start chat
Features
Overview
1.3.2 User Interface
Chat completions
1.2.1: show total tokens
1.3.6: Show error when request failed ( tested as rate limited )
You can use simply type and press
Enter
Context
1.0.0: context is valid until before popup close
1.1.4: maintain 6 messages in chrome storage
1.2.10: this feature move to configurable. default is false
Maintain Context
You can turn on Maintain Context
in the extension options to include up to 6 messages in your chat message. Maintaining context allows for a more natural conversation, but it uses more tokens.means: more costs
ℹ️ IMPORTANT:
CheapGPT maintains context by sending the previous 6 messages together. Please be aware that before use. Open AI API pricing is based on token count and model type.
ℹ️ IMPORTANT:
Chrome extension allow only 8k storage per field. so, if 6 messages is over the 8k. not saved from oldest message.
Conversations is maintained only 6 messages and saved in your chrome storage. It is keep the context of the chat while excluding the image generation history.
Select model
1.4.3: Add web search tools model (gpt-4o-search-preview and mini)
You can select a model
As more features or popular models become available in the future, the options may increase. For now, you can choose from one of three.
Multiline input
Multiline support
You can enter multiple lines by pressing shift + enter
.
Clipboard image (vision)
1.1.0: support clipboard image
You can use clipboard image
If you want to upload image from PC, just copy image file from explorer and paste to input.
ℹ️ IMPORTANT:
Image is need token too. please read Low or high fidelity image understanding first. Be aware of how much it will cost. CheapGPT is use default detail option "auto".
Pause and resume follow scroll
While the response message is appending, the scroll will automatically follow the new messages. If you scroll up, it will pause, and when you scroll back down, resume following the new messages.
Keyboard shortcut
1.1.5: Added toggle
You can use keyboard shortcut with Ctrl + Shift + X
on browser. It can be used faster and more conveniently.
1.2.2: Add clear message log
You can clear message log with Ctrl + Shift + L
. on message input
1.2.3: Add summary page
You can summary with Ctrl + Shift + F
. on message input
1.3.0: Add generate image
You can open dalle form with Ctrl + Shift + Y
. on message input
1.3.0: Add table of keyboard shortcuts
You can see keyboard shortcuts everytime in extension options page
Summarize Selection
1.2.0: Summarize Current page
1.2.3: Summarize into 3 key points
1.3.4: Show content length
1.3.4: Content sanitizing with DOMPurify
1.4.0: Change summarize target to selection text instead of page
1.4.6: Summary model is fixed to '4o-mini' when selected model is not flagship models
You can easily request a summary of the current selection texts into 3 key points. summary response is not used in context messages. in other word, not save.
Language configuration
1.2.4: Summarize language configuration
You can choose summary language in extension options
default is auto. detect in content.
Generate Image
1.3.0: support dalle image generation
1.3.6: Auto download image
open | input | response |
---|---|---|
![]() | ![]() | ![]() |
photos: generate image with dall-e
You can fill out the DALL-E form to request image generation. Messages are not retained, and the received images can be saved with a right-click menu.
Roadmap
- Simple chat completion
- Model select
- 1.1.0: Clipboard image
-
1.1.4: Maintain 6 messages in chrome storage1.2.10 - 1.2.0: Current page summarytext only
- 1.3.4: Content sanitizing with DOMPurifySummary page
- 1.2.10: Maintain context move to configurable
- 1.3.5: Add flagship model option 'o3-mini'
- Configurations
- 1.1.1: Select default model
- 1.2.4: Select summary language
- Keyboard shortcuts
- 1.1.5: toggle
- 1.2.2: clear messages
- 1.2.3: summary page
- 1.3.0: DALL·E 3 image generation request
- 1.4.3: Support 'Find on the web'
- 1.4.4: Download conversations