Your (temporary) Shareable Link will be valid for 7 days - for permenant links contact us at [email protected]
Here's what we care about the most:
Imagine if a super-smart person knew everything in the world. You could ask them anything, and their answers would depend on how good their info is. It's the same with AI. If it learns from bad info, it won't be very smart.
Just like people, an AI brain depends on the information it gets. If it only learns from one book or one TV show, news channel, or person, it won't know much about the world, or worse yet, it might have biases.
We use GPT models, some of the best AI brains in the business, for most of our cool tools. But we know there are other awesome AI brains out there you might not know yet. They're getting better all the time, and we're here for it!
We've got a buffet of AI brains for you! Our go-to is a speedy one called GPT-3.5-TURBO, and itβs super plugged in (like, real time internet-level plugged in). Want to explore? We've got more AI brains ready to go!
We let our friends (that's you!) pick the AI that's right for them. Whatever you're looking for, we've got an AI brain that can help. Dive into the world of AI with us and see what's possible!
(real time plugged in)
GPT-3.5-turbo is a state-of-the-art language model developed by OpenAI. It is capable of generating human-like text, understanding and responding to natural language prompts, and performing a wide range of language tasks, like chat, text-summarization, and more. This is our default go-to model.
(real time plugged in, and can take tons of data)
gpt-3.5-turbo-16k: allows for large amounts of text (hence the 16k) input and output, which we generally call 'context'. The previous model supports up to 4k context, which in our experience is more than enough for the most of the world. The 16k is available in our paid service.
(The Facebook (META) creation, they gave us a great gift)
Llama: LLaMA is a foundational large language model developed by Meta AI, Facebook's parent company. It ranges from 7 billion to 70 billion parameters (data) and is designed for research and commercial use. It is based on the transformer architecture and offers an open-source alternative to other LLM models. We provide 2k context, and believe this model is truly a contender to the GPT models in the future, it's getting better and better as time goes on. You can use this model for much of the common AI tasks.
(Vicuna - what a name!)
Vicuna: The Vicuna LLM model is a chat assistant developed by fine-tuning Llama 2 on user-shared conversations collected from ShareGPT. It achieves high quality marks in user preference tests and outperforms other models. The creators are: Large Model Systems Organization (LMSYS Org) which is an open research organization founded by students and faculty from UC Berkeley in collaboration with UCSD and CMU". Preliminary evaluation using GPT-4 as a judge shows Vicuna achieves more than 90%* quality of OpenAI ChatGPT
(Palm - I know what you are thinking, it's fast!)
Palm: The Palm Google T5 LLM model is a large language model developed by Google with 540 billion parameters. It is part of Google's Pathways Language Model (PaLM) project and is designed for tasks such as language understanding and generation. The model excels at tasks that require understanding ambiguous and figurative meanings of words. Even better, it does amazingly well of getting to the point with text summarization, though you do need to break up the document in 500 word chuncks. We like it - sometimes!
(This company is changing the game)
Mistral: Mistral LLM model is Mistral AI's first Large Language Model (LLM) with 7 billion parameters. Mistral-7B-v0.1 is a small, yet powerful model adaptable to many use-cases. Mistral 7B is better than Llama 2 13B on all benchmarks, has natural coding abilities, and 8k sequence length. We think these guys are going somewhere, they have made a lite model that many developers in the world can deploy easily.
(Not Parton and not DALL-E)
Dolly: Dolly LLM is the first open-source, instruction-following large language model (LLM) by Databricks. It has been fine-tuned on a human-generated instruction dataset for research and commercial use, boasting 12 billion parameters. Dolly allows businesses and developers to understand and execute complex instructions effectively. We like playing with this one sometimes, because it is a bit 'crazy' if you get our point, but in all fairness it does produce some good stuff.
(The iteration of GPT3.5, with some extras added)
GPT-4: GPT-4 LLM is a large multimodal language model that accepts text and image inputs and produces text outputs. It exhibits human-level performance on various professional and academic benchmarks, with improved reasoning and problem-solving abilities. We provide this model to our paid customers as it is the iteration from 3.5 and appears in certain circumstances to produce a higher quality output. Most people would not see the difference, unless you are in the AI everyday with complex tasks.
WEOPI Search harnesses the power of OpenAI's GPT to deliver precise and timely results. Here's how to get the best out of it:
Remember, the power of WEOPI Search lies in its ability to understand and adapt to your queries. The clearer you are, the better it serves. Happy searching!