Install
Add Prompt Paul to Chrome and keep AI assistance inside your current workflow.
Context-aware AI in the browser
Prompt Paul
Short feedback loops for selected text, without ceremony.
Get AI insight from selected text, page context, images, audio, and video without leaving your browser. Prompt Paul turns the moment you are already in into a quick, focused conversation.
Selected text actions
Page-aware side panel
Custom instructions
Tools and skills
Select
Highlight text, media, or page content and choose the action you need.
Respond
Receive concise analysis, proofreading, explanation, or custom instruction output.
Stay focused
No tab switching. No ceremony. Just a short loop from question to answer.
What it does
Built for fast context work.
Prompt Paul is designed around the small decisions that happen while reading, writing, debugging, researching, and drafting. It keeps the loop short and leaves the page where it belongs.
Selected text actions
Analyze, explain, proofread, summarize, or run a custom prompt against selected text from the context menu or side panel.
Page-aware chat
Load the current page into the side panel, ask follow-up questions, and keep the page context attached to the conversation.
Media-aware prompts
Work with images, video, and audio when your model supports multimodal input. Right click, ask, and get a response.
Save custom instructions
Store repeatable workflows for emails, meeting notes, captions, research summaries, code explanations, product descriptions, and support replies.
Tools and agent skills
Enable tool use for APIs, databases, search, code execution environments, and Skills Compatible agent skills when you need more than a plain chat response.
One-off instructions
Need a one-time response? Select text, enter a specific instruction, and get a tailored result without setting up a saved workflow.
Custom instruction guide
Adding a custom instruction to PromptPaul.
Let's give PromptPaul superpowers written just for you by you.
STEP 1 — Getting to the configuration page
Open up the configuration page of PromptPaul. To do this:
- Click the puzzle piece icon in the top right of your browser.
- A drop-down menu will appear with all your enabled extensions.
- Click the pushpin icon to the right of the extension you want to pin.
- Now simply click on the Prompt Paul icon and click “Configure Settings”.
STEP 2 — Writing the instruction
In the “Instruction Name” field, give your custom instruction a name such as
LinkedIn Outreach Message.
Next, type out what PromptPaul should do when you use this custom instruction. Using our example from above, we could type:
Create a personalized LinkedIn message to establish a connection and spark conversation. Ensure the message is:
- Personalised to the recipient's profile and interests
- Professional and polished
- Relevant and concise (1-2 paragraphs)
- Includes a clear call-to-action
Follow these guidelines:
- Formal greeting using first name
- Brief self-introduction (I am Prompt Paul a chrome extension that helps you leverage AI without breaking flow and with your context)
- Clearly stated purpose and value proposition (Stated Purpose PromptPaul simplifies and improves AI accessibly with instant relative and context aware output. Value Proposition Relative AI Outputs: Ai augmentation based on what your actually working on. Time-Saving: Remove's repeated prompting, saves effort with custom set instructions. Ease of Use: User-friendly for non-technical users. Customisation: You can tailor and build custom instructions for diverse use cases such as reaching out, quick business email replies, summarising texts etc. Enhanced Productivity. Specific call-to-action (Try out the Prompt Paul Chrome Extension)
- Professional digital signature (sign off as PromptPaul)STEP 3 — Saving the instruction
Click the save button. You should now find the custom instruction in your right-click context menu.
Workflow
Three moves. No ceremony.
Select text or page context
Use the context menu, right-click media, or load the current page into the side panel.
Choose an action
Pick a saved instruction, run a quick action, or ask a one-off question.
Get the loop back
Return to the page with a concise answer, proof, summary, or next step.
Compatible providers
Bring your model.
Prompt Paul works with OpenAI-compatible endpoints and popular routing or local model setups. Configure the endpoint, API key, and model that fit your workflow.
Google Gemini
OpenRouter
Groq
OpenAI
LM Studio
Together.ai
SambaNova
Provider setup guide
Now that you have chosen your provider.
It's time to set up the API endpoint, API key, and model context window. This is a once-off process, unless you want to try different models.
STEP 1 — Getting to the configuration page
Open up the configuration page of PromptPaul. To do this:
- Click the puzzle piece icon in the top right of your browser.
- A drop-down menu will appear with all your enabled extensions.
- Click the pushpin icon to the right of the extension you want to pin.
- Now simply click on the Prompt Paul icon and click “Configure Settings”.
STEP 2 — Setting the API endpoint
To set the model up for use, paste the LLM API endpoint URL for your provider.
Example endpoints as of 2026/06/21:
- Groq's API endpoint is:
https://api.groq.com/openai/v1/chat/completions - OpenAI's API endpoint is:
https://api.openai.com/v1/chat/completions - TogetherAI's API endpoint is:
https://api.together.xyz/v1/chat/completions - OpenRouter's API endpoint is:
https://openrouter.ai/api/v1/chat/completions - Google Gemini's endpoint is:
https://generativelanguage.googleapis.com/v1beta/openai/chat/completions - LM Studio's API endpoint is: This is dependent on your settings. Refer to the LM Studio documentation.
STEP 3 — Setting the API key
Paste or type in the API key you received from your provider into the input field.
STEP 4 — Setting the model name
Paste or type in the model name you wish to use.
Example model names:
- OpenAI:
gpt-5.5 - Groq:
llama-3.1-8b-instant - OpenRouter:
nvidia/nemotron-3-nano-omni-30b-a3b-reasoning:free - Google Gemini:
gemini-3.5-flash - TogetherAI:
zai-org/GLM-5.2 - LM Studio: This is dependent on your settings. Refer to the LM Studio documentation.
STEP 5 — Max tokens
You can simply leave the default of 2048. If you wish to consume longer texts, refer to your provider's documentation to find the token limit for your model.
STEP 6 — Saving the settings
Click the green “Save Settings” button.
API key guide
Get the key for your provider.
This guide covers API keys from OpenAI, Gemini, Groq, OpenRouter, Together AI, and LM Studio so you can connect Prompt Paul to the models you want to use.
Are there any free providers?
Yes. Groq, OpenRouter, and Google all offer multiple free options. LM Studio uses models running on your computer locally.
What is an API key and why do I need one?
An API key is a unique identifier that authenticates your requests to a service's API. Think of it as a password that allows Prompt Paul to communicate with the provider's servers.
1. OpenAI
Account Creation
Go to the OpenAI website and sign up for an account. If you already have an account, log in.
Accessing API keys
Once logged in, click your profile icon in the top right corner, then select “View API keys”. Alternatively, you may be redirected to the API section directly.
Click “Create new secret key”. A new API key will be generated. Copy it and store it somewhere safe. You will not be able to see it again.
Important notes
OpenAI may require billing information before using their API.
Always keep your API keys secure and do not share them.
2. Gemini (Google AI)
Account Creation
Visit the Google AI Studio website and sign in with your Google account.
Accessing API keys
After logging in, look for the “Get API key” button, usually located in the top right corner. Click “Create API key in new project”.
You will be prompted to create a new Google Cloud project or use an existing one. Follow the instructions. Once the project is set up, the API key will be generated and displayed. Copy and securely store it.
Important notes
Using Google AI APIs may incur costs. Review Google's pricing policies before making API calls.
Google Cloud Console might be needed to manage projects and API settings.
3. Groq
Account Creation
Go to the Groq website and create an account.
Accessing API keys
Navigate to the API section of your account after logging in. You should see an option to generate a new API key. Copy and store the generated API key securely.
Important notes
Groq's API access might be subject to usage quotas and pricing. Check their documentation for the latest details.
4. OpenRouter
Account Creation
Go to OpenRouter and sign up for an account.
Accessing API keys
After logging in, navigate to your account settings or API keys section. Click “Create API key” or a similar button. Copy the generated key and keep it safe.
Important notes
OpenRouter provides access to multiple LLMs through a single API. You need to set up billing to use the API.
OpenRouter charges for usage, but it usually costs less compared with other LLM API providers.
5. Together AI
Account Creation
Go to the Together AI platform and create an account.
Accessing API keys
Locate the API key section within your account after logging in. Click the “Create API Key” button. Copy the generated API key and save it securely.
Important notes
Together AI also offers billing plans. Check their website for details.
6. LM Studio
LM Studio is a local application, not a cloud service. It is designed to run models locally on your computer and does not require API keys.
Accessing models
Download and install LM Studio from its official website. After installation, you can download various models and use them locally.
Important notes
Because the models run locally, performance depends on the hardware of your machine.
General API key safety tips
- Never share your API keys publicly. Avoid posting them on social media or in public repositories.
- Use environment variables when possible instead of hard-coding keys directly into applications.
- Rotate keys periodically to improve security.
- Be aware of pricing. Some API providers charge for usage, so review pricing models to avoid unexpected costs.
Keep this guide as a starting point
This guide provides a general overview. Always refer to the specific documentation and instructions provided by each platform for the most accurate and up-to-date information.
Provider comparison
Understanding different AI API providers.
AI APIs act as the bridge between your applications and AI models, enabling features like text generation, image recognition, and more. The right provider depends on your project, budget, latency needs, and model preferences.
Providers covered
- OpenAI
- Gemini (Google AI) — offers free-use API keys.
- GROQ — offers free-use API keys.
- OpenRouter — offers free-use API keys.
- Together.ai
- LM Studio
- SambaNova — offers free-use API keys.
Choosing the right provider
The ideal AI API provider depends heavily on your specific project needs and constraints. Consider these factors before committing to one setup.
- Cost: How much are you willing to spend on API usage?
- Performance: What level of accuracy and speed is required?
- Ease of use: How important is integration quality and API documentation?
- Customization: Do you need to fine-tune or train your own models?
- Specific use cases: Which models are most suitable for your tasks, such as text, images, or code?
1. OpenAI
OpenAI is a leading AI research and deployment company known for its powerful language models, including the GPT series, and other AI tools like DALL-E for image generation.
Pros
- State-of-the-art models: Offers cutting-edge models that consistently perform at the top of benchmarks for various tasks.
- Wide range of capabilities: Provides models for text generation, code completion, image generation, and more.
- Extensive documentation and community: A large user base results in strong support and many online resources.
- Developer-friendly API: Well-documented and relatively easy to integrate into applications.
- Regular updates: Consistently releases model improvements, enhancing performance.
Cons
- Cost: Can be expensive, especially for high-volume or complex applications.
- Usage restrictions: Can sometimes have limitations on usage due to high demand or specific policy requirements.
- Black box nature: The inner workings of its models are often opaque, making it hard to debug specific issues or understand model bias.
2. Gemini (Google AI)
Gemini is Google's suite of AI models. Known for its multimodal capabilities, it handles text, code, audio, images, and video.
Pros
- Multimodal capabilities: Excels at handling and processing various types of data inputs and outputs.
- Integration with Google services: Easily integrates with other Google Cloud Platform offerings, like cloud storage.
- Scalable infrastructure: Benefits from Google's robust and scalable infrastructure, which can support high-demand applications.
- Competitive pricing: Offers various pricing tiers, which may be more competitive for some use cases than other leading providers.
Cons
- Newer in the market: Compared with OpenAI, its models are not as battle-tested or widely used by the community.
- Complexity: The wide variety of options can lead to a steeper learning curve for developers and a potential increase in setup time.
- Varying model performance: While some models excel in some areas, others may not be as good as others.
3. GROQ
GROQ is an AI accelerator company that focuses on super-fast AI inference through its Language Processing Unit (LPU).
Pros
- Incredibly fast inference speed: Boasts extremely fast inference speeds when compared with other platforms.
- Cost-effective: With faster inference speeds, there is the possibility to reduce cost at scale.
- Focus on speed: Optimized for high-throughput and low-latency AI inference.
- Streamlined API: Designed to minimize complexity and ease integration.
Cons
- Limited model selection: Currently focuses on supporting specific types of models which may not be appropriate for every use case.
- New platform: New in the market and may not have the community and third-party tool support other established providers have.
- Less mature: May have limitations in features that more mature platforms offer.
4. OpenRouter
OpenRouter is an API aggregator that provides a single access point to multiple AI models from various providers, including OpenAI, Google, Anthropic, and more.
Pros
- Unified access: Provides access to many AI models through one single API, simplifying workflows.
- Model flexibility: Allows users to easily switch between different models and providers.
- Cost comparison: Allows users to compare the cost of different models across various providers in one convenient place.
Cons
- Potential latency: May introduce slight latency due to the additional layer of aggregation.
- Dependency: Relies on the stability of the providers it connects to, introducing a potential point of failure.
- Feature variation: Features across different models may vary, requiring careful evaluation on a per-model basis.
5. Together.ai
Together.ai is a cloud platform for AI development, focused on open-source models, offering cloud APIs and custom model training and deployment.
Pros
- Open-source focus: Emphasizes open-source AI models, offering transparency and community contribution.
- Customization: Ability to fine-tune and deploy custom models for niche applications.
- Competitive pricing: Can be more affordable for specific workloads compared with commercial models.
- Strong community: Backed by an active community of AI researchers and developers.
Cons
- Model maturity: Some open-source models may not be as mature or perform as well as commercial alternatives.
- Complexity: Requires a deeper understanding of AI model training and deployment for custom models.
- Resource intensive: Custom training can require substantial computational resources.
6. LM Studio
LM Studio is a desktop application that allows you to run large language models locally on your computer, offering an alternative to cloud-based APIs.
Pros
- Privacy focused: All data and computations remain local to the user's computer.
- Offline capabilities: Ability to run AI models even without an internet connection.
- Model control: Offers the user complete control of the models being used.
Cons
- Resource intensive: Running large models locally requires a powerful computer with high processing power and sufficient RAM, and may not be available to some users.
- Limited scalability: Not suitable for high-volume applications that require scalable infrastructure.
- Initial setup: Requires the user to have the know-how to download, configure, and run models.
7. SambaNova
SambaNova offers AI infrastructure and model deployment options focused on enterprise-grade performance, reliability, and scalable AI workloads.
Pros
- Enterprise focus: Built for organizations that need reliable AI infrastructure and deployment support.
- Scalable workloads: Designed to support demanding AI applications across teams and production systems.
- Managed infrastructure: Can reduce the operational burden of running and scaling models.
Cons
- Enterprise fit: May be more appropriate for larger teams or organizations than casual individual users.
- Setup complexity: Enterprise AI platforms can require more planning and integration work.
- Pricing clarity: Costs may depend on deployment scale, support needs, and specific workload requirements.
How to choose
By carefully evaluating cost, performance, ease of use, customization needs, and your specific use case, you can select the right AI API to power your project.
Disclaimer
Pricing, capabilities, and specific features may change. It is always recommended to consult the official documentation of each provider for the most up-to-date information. This guide is for general informational purposes only.
Privacy
Local settings. Explicit sends.
Prompt Paul keeps configuration in Chrome storage and only sends content to your chosen LLM endpoint when you explicitly ask it to process something.
- API keys and settings stay in Chrome storage.
- Analytics are opt-in and disabled by default.
- No page URLs, page titles, page content, API keys, or prompt text are sent to analytics.
- Restricted browser pages are blocked before extraction.
View full privacy policy
At Prompt Paul, we are committed to protecting your privacy and ensuring transparency about how we handle your data. This Privacy Policy outlines the types of information we may collect, how we use it, and the steps we take to safeguard your information.
Information We Collect
We may collect the following types of information anonymously for usage statistics and to improve the functionality of the extension.
Location Information
We may collect general location data such as region, IP address, or approximate GPS coordinates, never your exact location. This information is used to understand the geographic distribution of our users and improve our services.
User Activity
We may collect data related to your interactions with the extension, such as clicks, mouse position, scroll activity, and keyboard activity. This information helps us analyze usage patterns and enhance the user experience.
How We Use Your Information
- To analyze usage statistics and improve the functionality of Prompt Paul.
- To understand user behavior and preferences to enhance the extension’s features.
- To ensure compliance with legal and regulatory requirements.
Data Collection Practices
All data collected is anonymous and cannot be used to personally identify you. We do not sell, trade, or share your information with third parties for marketing purposes. Data is aggregated and used solely for internal analysis and improvement of the extension.
Data Security
We implement industry-standard security measures to protect the data we collect. However, no method of transmission or storage is 100% secure, and we cannot guarantee absolute security.
Your Choices
You can disable data collection by adjusting the settings in the extension or uninstalling the extension. By using Prompt Paul, you consent to the collection and use of your information as described in this Privacy Policy.
Changes to This Privacy Policy
We may update this Privacy Policy from time to time. Any changes will be posted on this page with an updated “Last Updated” date.
Contact Us
If you have any questions or concerns about this Privacy Policy, please contact us via the Chrome Web Store listing page.
Thank you for using Prompt Paul.
Question Help & FAQ
Quick answers.
The short version: Prompt Paul is for people who want AI help close to the text they are already working with.
What does Prompt Paul send to the model?
Only the selected text, page content, media, or instruction you explicitly choose to process, plus your API configuration and prompt.
Where are API keys stored?
API keys are stored in Chrome storage on your device and are sent only when you send a prompt or run an action.
Can I use local or self-hosted models?
Yes, if they expose an OpenAI-compatible endpoint. LM Studio and similar local setups can work when configured correctly.
Can I save repeatable instructions?
Yes. Custom instructions can be saved and reused for common workflows like email replies, summaries, captions, research notes, and code explanations.
Does it support tools or agent skills?
Yes. Prompt Paul can route instructions through tools and Skills Compatible when enabled in your configuration.
What happens to analytics?
Analytics are opt-in. When enabled, they send coarse event labels such as instruction name, model name, and tool names. They do not send page content, URLs, titles, API keys, or prompt text.
Have a sharp edge to sand down?
Send feedback, bug reports, provider notes, or workflow ideas directly by email.