The development of conversational AI chatbot systems from competing companies in the summer of 2023 is intense, yet there’s no single system that excels at all tasks. For instance, Anthropic launched Claude.ai on July 11 to provide public access to Claude 2, which supports long documents and comes with enhanced content generation and coding skills. Two days later, Google expanded Bard’s availability to more countries and in more languages, along with the ability to modify responses and pin and share chats, among other improvements.
This article compares the core capabilities of Google Bard and Claude.ai, both of which are free and considered either experimental or in beta as of July 2023. To use Bard, you must sign in with a Google account. To use Claude, you must sign in either with a Google account or an email address. In either case, if you use a Google Workspace account, you may want to check with your administrator to make sure access is allowed.
Jump to:
Bard is an experimental conversational AI system designed by Google. As a large language model, Bard can generate responses of text, code and other content in response to an entered prompt or image. Google launched Bard in the first quarter of 2023.
Claude.ai is a public-facing beta of Anthropic’s conversational AI system. It was launched at the same time as when the company announced Claude 2 in mid-2023. Anthropic has emphasized its efforts to design Claude with an emphasis on AI safety. The company offers access to various editions of Claude (e.g., Claude 1, Claude Instant and Claude 2); Claude 2 drives Claude.ai interactions.
The chart below highlights the major differences between Google’s Bard and Anthropic’s Claude.
| Feature | Bard | Claude.ai |
|---|---|---|
| Made by | Anthropic | |
| Available countries | 230+ | U.S. and U.K. |
| Language support | 40+ | English. Some support for other languages, most notably Portuguese, French and German. |
| Coding support | 20+ languages | Yes |
| Internet access | Yes | No |
| Accepts image upload | Yes | No |
| Accepts text file upload | No | Yes |
| Context window* | Not specified. “Purposefully limited,” according to a Bard FAQ. | Approximately 75,000 words. |
| Export options | Gmail, Google Docs, Google Sheets, public link, copy content | Copy content |
* The context window determines the length of content the system can retain and respond to at once. Put simply, if you exceed the limit, the system may provide a response that indicates a lack of “retention” or “understanding” of content beyond the window. A long window makes the analysis of long documents possible.
Bard and Claude.ai are free to use. Neither Google nor Anthropic offers a paid edition of Bard or Claude.ai. The most notable competition to these two systems, ChatGPT from OpenAI, offers both a free edition and a more capable edition for an upgrade of $20 per month.
However, Google and Anthropic seek to serve enterprise and developer AI needs. Google offers a wide range of AI and machine learning systems, while Anthropic offers access to distinct editions of Claude, each optimized for different situations and price points.
Both Bard and Claude operate as conventional AI chatbots: You type a prompt, and the system replies with a response. As conversational chat systems, both Bard and Claude let you reference prior prompts in a chat. For example, if your first prompt is a request to summarize a book:
Can you summarize the key recommendations in the book, “Power and Progress: Our Thousand-Year Struggle Over Technology & Prosperity”?
A later query in the chat conversation with the intent to learn the authors might be:
Who wrote it?
The system should accurately understand that “it” refers to the book and “who” refers to the authors. Contrast that with a legacy search, where a summary of key points is simply not feasible and where you would need to combine everything into one keyword search query just to learn the authors.
The one significant difference between the two systems is that Claude supports a much longer context window than Bard. The result is that Claude can work with longer texts and also allows a more extended conversation about a topic than Bard.
Bard and Claude work well when you want the system to generate text. For example, you can ask either system to write text you need for pretty much any purpose, such as an email, marketing copy, a cover letter or a blog post. Specify in as much detail as possible the contents, length and type of text you want, and both systems will tend to produce usable text. Not happy with the result? Describe exactly how you need the tone, content, length or structure of the text revised, and ask Bard or Claude to try again.
Bard and Claude excel when you seek suggestions. Ask either chatbot for lists of things you want to learn about, explore or read, and you’ll likely receive at least a few relevant items. If you have examples, include them to provide context and nudge the system toward a more relevant response.
Remember, both of these systems may provide information that is inaccurate, misleading or simply wrong, so you may need to verify the content for accuracy.
SEE: Hiring kit: Prompt engineer (TechRepublic Premium)
Claude’s distinctive strength is its ability to work with long text files, thanks to the system’s significantly long context window. You may upload as many as five files of up to 10MB each along with your message.
For example, you might upload one or more sample technology plans, guides or strategy documents. Once shared, you could then explore various details within those documents with Claude (Figure A). Alternatively, you might provide Claude with relevant details about a different organization or client and ask the system to write a new plan, drawing from the prior example plans you had uploaded.
Figure A
Bard’s ability to access the internet and accept image input sets the system apart from Claude. Bard can leverage search to incorporate recent news, weather and other information into responses, unlike some large language models that lack access to information after a defined date. In many cases, this makes Bard able to provide up-to-date data in responses (Figure B), with a sample prompt about an event that occurred within the prior month. Additionally, since Bard incorporates many image capabilities previously launched in Google Lens, you may upload a JPEG, PNG or WebP file and then chat about the image content. (Bard’s image capabilities are available initially only when used in English.)
Figure B
Bard and Claude represent state-of-the-art conversational AI systems with rapidly evolving capabilities. Either system is an excellent option when you want to experiment with a generative AI tool.
However, Bard is clearly the best when your chat requires internet access or image support, while Claude is clearly the better option when your chat might benefit from uploaded documents or long conversational context.
Experiment with both AI systems to become familiar with their capabilities, and then select the chatbot that is most able to help you accomplish a specific task.
This comparison relied on detailed reviews of announcements from each vendor, experimentation with each system and manual testing of specific features publicly available as of mid-July 2023. Since both companies are adding capabilities fairly often, you may want to check the latest Google Bard updates or Anthropic announcements to learn of more recently added features.
