Classifies prompts sent to an LLM by task type (open QA, code generation, summarization, rewrite, ...) and scores their complexity (creativity, reasoning, knowledge, constraints) in 100+ languages with multilingual-prompt-task-complexity-small (Apache-2.0), running entirely in your browser. One prompt per line. Useful for routing prompts to cheaper or stronger models.
Loading model…
Labels follow an LLM teacher's judgement and complexity is subjective; validate thresholds on your own traffic. See the model card for measured accuracy.