When it comes to learning a second language—especially if you're trying to learn Spanish with English as your primary language—large language models can be incredibly helpful. While they may not always be the most reliable sources for factual information, they excel at facilitating discussion and identifying logical errors in speech.
In my experience, Grok 3 and Grok 4 (including the "Heavy" version) tend to perform better when it comes to English-language processing and factual accuracy. However, when it comes to handling multiple languages simultaneously—like asking questions in English to learn Spanish grammar or sentence structure in Spanish—they don’t seem to handle it as effectively. Based on professional testing, Grok 3 still falls short in this area, and there's currently no substantial data available on Grok 4's performance in multilingual settings.
On the other hand, ChatGPT has had more time in the public space, and it shows. So far, it's been the most effective tool I've used for multilingual discussions, particularly between English and Spanish. Whenever I want to dive into a conversation in Spanish, I find myself reaching for ChatGPT instead of Grok.
To be fair, I haven’t ventured much beyond these two models for bilingual conversations. I might experiment with Google’s Gemini at some point, but for now, ChatGPT remains my top choice. When it comes to Spanish-only discussions, there are models that are recommended that are funded by the Spanish government, but I haven't tried them yet. They're called MarIA and ALIA.