Qwen: Qwen3 Coder Next

# 1. Pull and execute model locally with Ollama
ollama run qwen-qwen3-coder-next

# 2. Or invoke via local OpenAI-compatible endpoint
curl http://localhost:11434/v1/chat/completions \
  -H "Content-Type: application/json" \
  -d '{
    "model": "qwen-qwen3-coder-next",
    "messages": [{"role": "user", "content": "Analyze reasoning chains on AIME 2026"}]
  }'