4 Best Free AI Search Engines

Crazy news guys, we’ve just launched a startup program for AI founders (the perks are crazy).

You can get $2000 worth of credits (Anthropic, Mistral, OpenAI, and Phospho) + a call with our amazing team to guide you in your product-market-fit journey.

You can apply here.

Traditional search engines as we know them today produce results to queries by relying on algorithms that combine different keywords, how often a page is referenced, content quality, how fast the page loads, amongst other things.

These factors determine what the search engine thinks will be the most helpful response. However, they often struggles to understand natural language queries or the real intent behind searches, so there’s difficulty in handling more ambiguous queries. It’s also unable to produce personalised responses to us which can make finding answers really cumbersome.

The problem is, and I never thought we’d reach this point, but googling things can honestly feel like a chore. Not only do we have to hop through several links to actually find what we're searching for, but we also navigate a maze of ads and pop-ups. Even through all this we often don’t find what we need so why not see what AI search can offer us.

An AI search engine still uses this basic system, but instead of showing you a list of pages, it tries to answer your question itself in a conversational way after crawling through all the pages for you to provide summaries.

More importantly, you can now get answers without being sent down rabbit holes clicking through every link manually just to find what you want. It's as if you have a personal search engine assistant.

It passes all the information it collects through the AI model it uses, which then summarises it in a couple of paragraphs whilst (usually) citing its sources. AI's is able to understand natural language meaning it can ‘get’ your questions really well, even if you don't use keywords that traditional search would need.

All sorts of companies ranging from huge players like Microsoft to smaller startups are offering us AI search engines with the ability to generate better answers to our queries and faster.

In this article, we’ll look over the top 3 best AI search engines that are free to use.

1) Perplexity

Unlike Google or Bing where AI was integrated into their existing search engines, Perplexity is built to be an AI first search engine which shows in its approach to the UI. It provides a nice balance of AI chatbot with traditional search engine. You have your chat field at the bottom to give it prompts but you’re also given suggestions based on current events.

Once you enter your search query, the results page provides several links at the top, which is helpful if you use Perplexity as a search engine to find the most appropriate website. The results include conversational, concise, bulleted AI-generated answers with footnotes and website links as well.

Another handy feature of Perplexity's conversational layout is that you can browse the threads of your previous searches and even save collections of pages if you want to organise your searches a bit more.

Lastly, underneath the AI insights, users can take advantage of a "related" section, which resembles the experience on a search engine tool, such as Google's "people also ask" feature. This "related" section encourages discovery.

This similarity with traditional search engines while providing cited summaries are what set Perplexity apart.

Perplexity is free to use even without an account, which makes it easy to jump in and try it out straight away. But if you want any premium perks, like document upload and more powerful AI models like GPT-4, you'll have to pay $20/month.

2) Microsoft Bing Co-Pilot

Microsoft added an AI chatbot to its Bing search engine ’Copilot’, and since then they’ve seen its daily active users on Bing surge with over 40 million new users during the past year alone.

Microsoft collaborated with OpenAI when developing Copilot, so the chatbot uses GPT-4 which is a top of the range LLM at the moment.

If you’ve tried it already you’ll know that Copilot in Bing has three conversation styles: Creative, Precise, and Balanced, which allows you to choose the type of output from the chat depending on what you’re using it for. This is quite good as it sidesteps any prompting you would otherwise have to do in order to get a specific type of response from your AI search engine.

The best part of Copilot’s UI is that the AI insights aren't overwhelming or limited to one box off to the side. This makes for a smoother UX as you can easily expand the results to see more AI insights. However, if you want to ignore the output and scroll through the normal search results, the AI insights are confined to a small portion of your desktop screen that you can ignore. If you aren't entirely sure if you want to commit to an AI search engine, Bing Copilot is a great option.

Naturally, there is still some room for improvement, but it’s an improvement over older AI search engines and is constantly getting better. The service is available to everyone free of charge, but you have to be signed in to your Microsoft account to get more than four query results per day.

3) Google Overviews

If you are a loyal Google user and curious about how AI can improve the search experience, then Google might already be the best option for you as it offers the best of both worlds.

Basically, when you type a search query into Google that could be optimised by AI overviews, Google will automatically show you the AI insights. The only downside is that if you are not interested in seeing them, you are automatically exposed to them. Since the AI insights are not as compact as Bing's and take up the top of the screen, pushing organic search results down may not be an approach that works for everyone.

The only other downside to Google's AI Overviews is it doesn't link sources to individual pieces of information like Perplexity does. Instead, it accumulates all the sources together at the end of a paragraph, so it's cumbersome to figure out where a particular fact came from.

However, if you do prefer Perplexity's conversational interface, you can also access Google's AI search capabilities from its Gemini chatbot. In here you can also prompt Gemini to create AI-generated images, describe or research your uploaded pictures for insights, and plug into other Google services. For example, you could connect it to your Gmail to ask for recommendations for an upcoming trip based on your itinerary.

4) Tak

Tak is a personalised AI search engine we’ve built here at Phospho which puts accuracy, relevance and safety at the forefront, and we do this in very interesting ways to provide a different option to the search engines mentioned above.

Tak is much like Perplexity in keeping the similarities of traditional search engines for better usability and familiarity, but the one thing it offers that Perplexity cannot is continuous learning and adaptation for more personalised and streamlined experiences over time. Tak constantly learns from user interactions to improve its algorithms, accuracy and contextual relevance over time through with the use of behaviour analytics.

For example, even when a search query is naturally difficult to understand, Tak will provide you with tick boxes to determine the right search intent behind your queries to provide more accurate results for similar searches in the future.

Tak also learns and improves on its own in the backend through knowledge graphs to map connections between different pieces of information. This helps to provide more interconnected and contextually consistent search results without constant manual fine tuning to get what you want.

Here’s a non exhaustive list of Tak’s core features:

• Conversational interface with up to date web data from continuously crawling and indexing web pages.
• Constant optimisation and self learning through user interactions to understand you better and provide the answers you’re actually looking for faster.
• Robust security around user privacy and data protection. We’ve ensured that search data is handled extremely securely and adheres to industry standards and regulations.

You can try out Tak for yourself here, it’s completely free to use!

Beyond general usage, the real game changer comes from combining Tak with our open source text analytics platform Phospho when building an LLM app to enable constant data driven iteration based on in-app user interactions.

What better way to fuel iteration that the feedback from in-app conversations your users are having with Tak. It integrates very easily with existing tools and popular tech stacks (e.g JS, Python, CSV).

Try integrating them both to see for yourself, sign up to use Phospho here and start leveraging usage data to streamline your iteration cycles.

The future of search with AI

As the internet's knowledge graph continues to expand, it will soon be impractical for us to search through it ourselves. People are always trying to find the fastest and easiest way to get answers.

With traditional search engines you often get a lot of links but no clear answer. AI search engines offer a quicker route to obtaining answers to queries by allowing you to converse with the search results instead, and pick out relevant links whenever you wish to dive deeper. The quality of results are usually better as well because they consider the context and actual meaning behind the question, rather than just keywords and phrases.

It will only become more and more important to test these different AI search engines for yourself to explore and determine which one best suits your needs. AI search engines will also learn from user behaviors and preferences to personalise search results for you so it will only improve with more usage.

To familiarise yourself with their advanced capabilities, try out Tak our AI search engine, or watch our 40 second demo video on YouTube here.