kras99 - stock.adobe.com
Microsoft Copilot vs. Google Gemini: How do they compare?
Microsoft Copilot and Google Gemini are frontrunners in the generative AI productivity tools space. Explore how they compare in terms of key features, capabilities and pricing.
Businesses are showing increasing interest in generative AI productivity tools like Microsoft Copilot and Google Gemini. But choosing the best option for a specific organization can be difficult.
For companies considering adopting a tool from one of these tech giants, a thorough evaluation is essential. See how Microsoft Copilot and Google Gemini stack up in terms of features, pricing, performance and integration with their respective ecosystems.
Microsoft Copilot vs. Google Gemini: Core features
After the release of ChatGPT in November 2022, Microsoft previewed Copilot -- initially Bing Chat -- in February 2023 and released it to the general public in May 2023. Thanks to Microsoft's strategic partnership with OpenAI, Copilot uses the same large language model (LLM) as ChatGPT, while integrating Bing Search for real-time information access.
Google, however, entered the race even earlier with the release of Bard in February 2023, rebranded a year later as Gemini. Throughout 2024, Google has made significant improvements to its language models.
Compare the following main features for each model:
- Context size. A model's context window describes how much information it can process at once -- essentially, acting as the model's memory. Developers using the Gemini API have access to a context window of up to 2 million tokens, while Gemini Advanced for end users can handle up to 1 million. In comparison, Copilot's LLM -- OpenAI's GPT-4 -- supports 128,000 tokens.
- Integration. Gemini integrates with Google Search, which is much more widely used than Microsoft's Bing. Gemini's overall integration support outside Google's ecosystem is limited; currently, Gemini integrates only with Google Workspace, and Copilot similarly integrates with Microsoft apps. Plans to extend Copilot via OpenAI functions, such as the OpenAI o1 model's advanced reasoning capabilities, could offer broader integration potential down the line.
- Multimodality. Both options offer image generation capabilities. Gemini uses Imagen 3, recently made generally available, while Copilot uses OpenAI's Dall-E 3.
- Performance. Gemini 1.5 Pro currently outperforms Copilot on needle in a haystack (NIHS) benchmarks that test how well an LLM can retrieve and use small, hard-to-find pieces of information from a large pool of text. But these benchmarks can quickly become outdated due to the rapid pace of model development.
- Plugins. Both Gemini and Copilot offer several plugin options. Gemini can connect to Google applications like YouTube, Google Maps and Google Hotels, while Copilot integrates with external services including Klarna, Suno, OpenTable and Shop.
Table 1
Comparison of Microsoft Copilot and Google Gemini's core features
| Feature | Microsoft Copilot | Google Gemini |
| LLM |
OpenAI GPT-4 |
Gemini 1.5 Pro |
| Price |
$20 per user, per month for Copilot Pro; $30 per user, per month for Microsoft 365 Copilot |
$19.99 per user, per month for Gemini Advanced after a 1-month free trial |
| Context size |
128,000 tokens, which is approximately 100,000 words |
1 to 2 million tokens, which is approximately 750,000 to 1.5 million words |
| Internet search integration |
Bing Search |
Google Search |
| API support |
No |
No |
| Customizability |
Yes, Copilot AI Agents; part of Copilot Studio, aimed at businesses |
Yes, Gems, which is similar to OpenAI's custom GPTs |
| Integration with collaboration suite |
Yes, integrated in Office apps with Pro |
Yes, integrated with Google Workspace with Gemini Advanced |
| Language support |
25 languages, with dialects for Chinese, English, French, Portuguese and Spanish |
46 languages |
| Text-to-image |
Yes, Dall-E 3 |
Yes, Imagen 3 |
| Text-to-voice |
Yes |
Yes, 10 different voices |
| Voice-to-text |
Yes |
Yes |
| Benchmarks (MMLU) |
88.4 (GPT-4o) |
81.9 (Gemini 1.5 Pro) |
Where does Amazon Q fit in?
Microsoft and Google aren't the only players in the generative AI space. Amazon has introduced its own service, called Amazon Q, with a Business Pro subscription priced at $20 per user, per month and a lite version with basic functionality for $3 per user, per month.
Amazon Q shares many features with Gemini and Copilot, but it stands out for its more extensive ecosystem of integrations and plugins and the ability to build custom integrations. Unlike Gemini and Copilot, which are primarily aimed at individual users and business productivity, Amazon Q is tailored more for businesses that use AWS, using services like AWS identity management for access and security.
Features and products on the horizon
Google has announced several new features for Gemini, including a real-time assistant with video and audio capabilities, thanks to its ability to handle large context sizes and video content. Google is also working on new application extensions and the option to integrate its LLM into the Android ecosystem, which should result in more localized language models running directly on Android devices.
Microsoft is taking a similar approach, embedding Copilot into its Windows OS and offering it as a standalone mobile app. Other recent updates include improvements to Microsoft 365 Copilot and Copilot Pages, a platform for collaborative sharing and editing of AI-generated content.
Generative AI tools for coding
Microsoft, Google and Amazon each offer generative AI tools specifically for software development. Through its partnership with GitHub, Microsoft provides GitHub Copilot, which is widely recognized for its code suggestions, completions and generation across multiple programming languages. The tool was also recognized in August 2024 as a leader in Gartner's first-ever Magic Quadrant for AI code assistants.
Google offers Gemini Code Assist, which uses Google's language models to provide real-time code suggestions and completions. Unlike GitHub Copilot, Code Assist remains under development, with some features still in preview, and lacks an extensive developer ecosystem. Amazon unveiled its coding assistant tool, Amazon Q Developer, earlier this year.
In terms of monthly pricing, GitHub Copilot is $10 per user, per month and Gemini Code Assist is an up-front annual commitment of $19 per user, per month, with a free trial option until November 8, 2024. Amazon Q Developer offers a free tier as well as a Pro tier for $19 per user, per month.
Copilot vs. Gemini: Which is best?
Choosing between these AI tools is challenging in part because generative AI tools are continuously evolving. In addition, LLMs are constantly improving, both in performance and context size, and vendors regularly release new features to remain competitive.
Ultimately, whether Copilot or Gemini is best will depend on which ecosystem a company uses. For companies primarily working with Microsoft applications and platforms, Copilot is likely the best option. But for teams invested in Google's suite of products, Gemini is typically the better fit.
Marius Sandbu is a cloud evangelist for Sopra Steria in Norway who mainly focuses on end-user computing and cloud-native technology.
