Spanish Future Tense Dataset for Evaluating LLM Choice: Morphological vs. Periphrastic

The dataset contains test questions to evaluate LLMs in Spanish 100 questions are about Spanish futures with prospective meaning. With two possible options (morphological future and periphrastic future), the LLMs must choose one of the two, both of which are possible in Spanish. 65 questions are about Spanish futures with epistemic meaning. With two possible options (morphological future and periphrastic future), the LLMs must choose the correct answer, with only the morphological future being possible in these contexts. Some sentences were drawn from the Corpus de Referencia del Español Actual (CREA) of the Real Academia Española and subsequently adapted for inclusion in the dataset. An example .xlsx file is provided with the results of 17 models from different companies: four models from Google (Gemini-2.5-flash, Gemma-3-27b-it, Gemma-3-12b-it, and Gemma-3-4b-it), three models from Anthropic (Claude-3.5-haiku, Claude-opus-4, and Claude-sonnet-4), one model from Meta (Llama-3.3-70b-instruct), one model from DeepSeek (DeepSeek-rl-0528), one model from Mistral (Mistral-nemo), three models from Qwen (Qwen-3-32b, Qwen-3-14b, and Qwen-3-8b), and four models from OpenAI (GPT-4o, GPT-4.1, GPT-4.1-mini, and GPT-4.1-nano).