Claude Instant with 100,000 tokens surpasses the main generative AI chatbots

Anthropic has released a new version of Claude that takes 100,000 tokens, or about 75,000 words of input.

This allows users to analyze and perform tasks on long excerpts of books, codes, documents, transcripts and more.

claude-instant-100k detailsScreenshot from Poe, June 2023

Read on to learn more about the latest version of the Anthropic AI chatbot, Claude-instant-100k, and see how it performs tasks against its competitors: ChatGPT, Bing AI, and Google Bard.

AI chatbot input limitations

Want to ask questions about content? Here’s the length of input you can send to AI chatbots from OpenAI, Microsoft, and Google.

  • OpenAI’s standard GPT-4 model offers 8,000 context tokens, while the extended context-length model offers 32,000 tokens/25,000 input words.
  • The input box for the Bing AI chat prompt shows 0/4000, suggesting it only accepts up to 4,000 characters. However, the chatbot accepted 40,400 lorem ipsum characters and accurately calculated that it was a total of 6,000 words.
  • Google Bard accepts up to 10,000 characters.

Claude-instant-100k accepts: three times more input than the best model GPT-4.

But how well can they follow directions?

I experimented with three activities using Claude-instant-100k, ChatGPT Plus, Bing AI and Google Bard. Here are the results.

Activity no. 1: Edit a 1,580-word list with Claude, ChatGPT, Bing, and Bard

The input: a 126-item list (some of the early ChatGPT plugins) at 1,850 words/12,500 characters.

The prompt: You can reword and edit the descriptions for the following list:

[the list]

Claude completed the task

  • Claude finished editing my list at item #126 without hallucinating and followed the prompts to reword the item descriptions.
  • Compared to other AI chatbots, Claude-instant-100k completed the task faster.

ChatGPT Default GPT-4 Hallucinated (Link to chat)

  • ChatGPT stops midway through item #. 37.
  • I clicked continue generating – ChatGPT finished number 37 and stopped again.
  • I asked to continue.
  • ChatGPT starts listing items that weren’t in my original input.
  • After a few more clicks to continue generating, ChatGPT finished editing my list to item #1. 142.

Tried Bing AI

  • Bing accepted the entire prompt and listing.
  • He asked if that was what I was looking for – I said yes and please continue.
  • After asking it to continue further, Bing finished changing my list to item #1. 126 without hallucinations.
  • Unfortunately, it stopped rephrasing descriptions after item #1. 18 and produced the list as I entered it.

Bard failed to create images

  • Due to the 10,000 character input limit, I could only submit 99 items from the list in the prompt.
  • Bard replied, “I can’t create images yet, so I’m unable to help you.” or “It’s not something I’m capable of doing yet.”

Finally, Claude-instant-100k completed the task with maximum accuracy and speed.

You can compare the output of Claude, ChatGPT and Bing with the original list of items in this Google Sheet.

Activity no. 2: Summarize a 23,500-word PDF with Claude, ChatGPT, Bing, and Bard

The input: 23,508 words / 125,104 characters copied from a PDF documenting TikTok’s responses to a hearing on how Congress can protect US citizens from the social platform.

The prompt: Please summarize the following:

[the text]

Claude summarized the entire text succinctly

claude-instant-100k summarizes the pdfScreenshot from Poe, June 2023

ChatGPT Default GPT-4 requested something shorter

Claude Instant with 100,000 tokens surpasses the main generative AI chatbotsScreenshot from ChatGPT, June 2023

Bing AI lost connection

Bing AI lost connectionScreenshot from Bing, June 2023

Bard summarized the first 10,000 characters

Google Bard summarizes pdfScreenshot from Google Bard, June 2023

Activity no. 3: Modify this PHP code with Claude, ChatGPT, Bing and Bard

The input: a PHP file with 14,787 characters.

The prompt: You can debug the following:

[the code]

Claude offered some suggestions

claude-instant-100k debugs the codeScreenshot from Poe, June 2023

The GPT-4 ChatGPT code interpreter determined that it was valid

ChatGPT code interpreterScreenshot from ChatGPT, June 2023

Bing AI could not debug your code

Bing Ai does not debug your codeScreenshot from Bing, June 2023

Bard deemed the first 10,000 characters error-free

Google Bard debugs your codeScreenshot from Google Bard, June 2023

Claude-instant-100k outperforms competitors

In a few initial experiments with everyday tasks, it’s clear that Claude’s latest release from Anthropic can outperform its competitors.

Like ChatGPT, experimental features for Claude come at a price.

Poe subscribers can Claude-instant-100k on a desktop browser for $19.99 a month, comparable to the price for ChatGPT+ subscribers.

Benefits for Poe subscribers Screenshot from Poe, June 2023

If you have tasks that require long inputs, give Claude’s new version a try.


Featured Image: Koshiro K/Shutterstock


#Claude #Instant #tokens #surpasses #main #generative #chatbots

Leave a Comment