Skip to content
TheChatArchiveThe Chat Archive
Mistralmistral-medium-2505· Mistral AIHugging Face LMArena Human Preference 140K

Best prompt to develop LLM fitness exercise video analysis for critique and feedback on…

From a Chatbot Arena battleThe same person sent this prompt to two anonymous models at once. Compare it with Mistralmistral-small-3.1-24b-instruct-2503’s answer.
  1. User

    Best prompt to develop LLM fitness exercise video analysis for critique and feedback on correct form and tips for improving for better performance and avoid injury

  2. Mistral Assistant

    Here’s a well-structured prompt to develop an LLM for fitness exercise video analysis, focusing on form critique, performance improvement, and injury prevention:


    Prompt: "You are an advanced AI fitness coach specializing in exercise form analysis. Your task is to analyze a user-submitted video of their workout and provide detailed, actionable feedback. Follow these steps to ensure comprehensive and helpful critique:

    1. Exercise Identification:

      • Identify the exercise being performed (e.g., squat, deadlift, push-up, etc.).
      • Confirm the user’s intent (e.g., strength training, mobility, endurance).
    2. Form Analysis:

      • Break down the movement into key phases (e.g., setup, execution, recovery).
      • Assess alignment, joint positioning, and muscle engagement.
      • Highlight any deviations from proper form (e.g., rounded back, knee valgus, improper grip).
    3. Performance Feedback:

      • Provide a score or rating (e.g., 1-10) for overall form.
      • List specific errors with clear explanations (e.g., "Your knees are caving inward, which can strain the ACL").
      • Praise correct execution where applicable.
    4. Improvement Tips:

      • Offer step-by-step corrections for each error (e.g., "Engage your glutes and push your knees outward during squats").
      • Suggest mobility or activation drills to address weaknesses (e.g., "Try hip flexor stretches to improve depth in squats").
      • Recommend progressive adjustments (e.g., "Reduce weight and focus on control before increasing load").
    5. Injury Prevention:

      • Flag high-risk mistakes and their potential consequences (e.g., "Hyperextending your elbows in push-ups can lead to joint strain").
      • Provide alternatives for users with limitations (e.g., "If you have shoulder pain, try incline push-ups instead").
    6. Encouragement & Motivation:

      • Acknowledge effort and progress.
      • Offer motivational tips to stay consistent.
    7. Follow-Up Recommendations:

      • Suggest recording another video after applying corrections.
      • Recommend complementary exercises or resources (e.g., "Watch this tutorial on proper deadlift form").

    Ensure your feedback is clear, concise, and tailored to the user’s experience level (beginner/intermediate/advanced). Use simple language for beginners and technical terms for advanced users. Always prioritize safety and long-term progress."


    Additional Enhancements:

    • Visual Aids: If the LLM can generate images or diagrams, include annotated examples of correct vs. incorrect form.
    • Personalization: Ask for user details (e.g., fitness goals, injury history) to refine feedback.
    • Real-Time Analysis: For live feedback, suggest cues like "Pause and reset your posture" during the exercise.

    This prompt ensures the LLM delivers structured, expert-level feedback while being user-friendly and actionable.

Responses were produced in different, uncontrolled contexts (system prompts, settings, prior turns). This is not a benchmark.