Mistral AI Challenges OpenAI With New GPT-4 Rival and Chat Assistant
Paris-based startup Mistral AI introduces Mistral Large, a flagship language model designed to compete directly with GPT-4 and Claude 2. The company also launches Le Chat, a beta chat assistant, as it shifts toward a commercial API business model.
Paris-based startup Mistral AI is making a major push to compete with OpenAI and Anthropic by launching a new flagship large language model called Mistral Large. This new model is designed to rival top-tier models like GPT-4 and Claude 2 in reasoning capabilities. Alongside the model, the company is introducing Le Chat, its own conversational chat assistant that is currently available in beta.
Although Mistral AI originally gains attention for its open-source focus and massive fundraising efforts, the company is shifting its business strategy to mirror OpenAI's approach. Unlike its earlier models, Mistral Large is not open source and is instead accessible exclusively through a paid API. The pricing is set at $8 per million input tokens and $24 per million output tokens, making it roughly 20% cheaper than GPT-4 Turbo while supporting a 32k-token context window.
Mistral AI claims that Mistral Large ranks second only to GPT-4 across several standard benchmarks, though independent testers still need to verify these results in real-world scenarios. The model currently supports multiple languages, including English, French, Spanish, German, and Italian. As the AI landscape evolves rapidly, this release solidifies Mistral AI's position as a formidable contender in the race to build foundational artificial intelligence technologies.