Performance Benchmarks for trinity-large-preview

Analyze how trinity-large-preview stacks up against other frontier models in terms of cost-to-performance and raw intelligence (LMArena Elo).

2026 Performance Summary

🚀 Intelligence (Elo): 1379
💰 API Cost (Input): $2.5e-7 / 1M tokens
⚡ API Cost (Output): $8e-7 / 1M tokens
⏱️ Real-world Throughput: 263.5 tokens/s

Analysis: While trinity-large-preview is a capable model, it is currently outperformed in efficiency by qwen3.7-max-preview. qwen3.7-max-preview sits higher on the Pareto line, offering approximately 96 points more intelligence (Elo) for a comparable or lower API cost.

Highly Recommended Alternatives

Frontier Model	LMArena Elo	API Cost (1M)	Throughput
qwen3.7-max-preview	1475	$0.00000125	44
gemini-3-flash	1473	$5e-7	79
mimo-v2.5-pro	1466	$4.35e-7	61
qwen3.7-plus	1463	$3.2e-7	11
gemma-4-31b	1451	$1.2e-7	153

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