model comparison · 248 golden queries

LLM + embed + rerankp50 latcost / qquality

gpt-4o-mini + e5-large + bge1.8s$0.0400.92current

gpt-4o + e5-large + bge2.4s$0.1800.95exp

gpt-4o-mini + bge-base + bge1.4s$0.0300.89exp

claude-sonnet + e5-large + bge2.6s$0.1200.94exp

llama-3.1-70b + e5-large + bge3.2s$0.0200.88exp

latency · 30일 추이

quality · 30일 추이

cost / day