Performance Benchmarks for gpt-oss-20b

Analyze how gpt-oss-20b stacks up against other frontier models in terms of cost-to-performance and raw intelligence (LMArena Elo).

2026 Performance Summary

🚀 Intelligence (Elo): 1318
💰 API Cost (Input): $3e-8 / 1M tokens
⚡ API Cost (Output): $1.3e-7 / 1M tokens
⏱️ Real-world Throughput: 367 tokens/s

Analysis: While gpt-oss-20b is a capable model, it is currently outperformed in efficiency by gemini-3-flash. gemini-3-flash sits higher on the Pareto line, offering approximately 155 points more intelligence (Elo) for a comparable or lower API cost.

Highly Recommended Alternatives

Frontier Model	LMArena Elo	API Cost (1M)	Throughput
gemini-3-flash	1473	$5e-7	72
mimo-v2.5-pro	1468	$4.35e-7	57
deepseek-v4-pro	1458	$4.13772e-7	59
hy3	1457	$1.32e-7	95
gemma-4-31b	1451	$1e-7	183

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