Qwen 2.5 72B Instruct (Alibaba Qwen) AI Model: Benchmark Scores, Pricing & VRAM Specs | ModelAnalysis.si

Alibaba Qwen • OPEN WEIGHTSQwen 2.5 72B InstructAlibaba premier open-weights foundation model excelling in multilingual reasoning, coding, and mathematical logic.

            Metric Dimension
            Verified Value
            Evaluation Benchmark
          
            Chatbot Arena Elo
            1290 pts
            LMSYS Live Battles
          
            Coding Index
            86.5 / 100
            SWE-bench Verified
          
            Agentic Tool-Use
            85 / 100
            Multi-Turn Tool Calls
          
            Input Rate
            $0.35 / 1M
            Commercial Token API
          
            Output Rate
            $0.40 / 1M
            Generation Billing
          
            Context Horizon
            128k
            Maximum Token Window
          
      Dedicated Cloud DeploymentDeploy Qwen 2.5 72B Instruct on high-performance dedicated NVIDIA GPUs at 30–60% off hyperscaler rates.
Deploy on RunPod Spot Cloud →