Performance Benchmarks for qwen2.5-vl-72b-instruct

Analyze how qwen2.5-vl-72b-instruct stacks up against other frontier models in terms of cost-to-performance and raw intelligence (LMArena Elo).

2026 Performance Summary

🚀 Intelligence (Elo): 1122
💰 API Cost (Input): $2.5e-7 / 1M tokens
⚡ API Cost (Output): $7.5e-7 / 1M tokens
⏱️ Real-world Throughput: 25 tokens/s

Analysis: While qwen2.5-vl-72b-instruct is a capable model, it is currently outperformed in efficiency by gemini-3.5-flash. gemini-3.5-flash sits higher on the Pareto line, offering approximately 358 points more intelligence (Elo) for a comparable or lower API cost.

Highly Recommended Alternatives

Frontier Model	LMArena Elo	API Cost (1M)	Throughput
gemini-3.5-flash	1480	$0.0000015	118
gpt-5.5	1478	$0.00000125	48.5
gemini-3-flash	1473	$5e-7	72
deepseek-v4-pro-thinking	1461	$4.35e-7	25
gemma-4-31b	1451	$1.2e-7	23

*These models represent the Pareto Frontier (optimal cost-to-performance).*