Cost Efficiency Leaderboard
Best score per dollar — which models give you the most Elasticsearch expertise for your money?
| # | Model | Score | Cost | Score / $ | Tokens |
|---|---|---|---|---|---|
| 1 | gpt-oss-120bopenai | 74% | $0.32c | 23,017 | |
| 2 | gpt-5.4-miniopenai | 93% | $0.032 | 2,927 | |
| 3 | gemini-3.1-flash-lite-previewgoogle | 96% | $0.113 | 852 | |
| 4 | gpt-5.4openai | 90% | $0.114 | 790 | |
| 5 | claude-sonnet-4.6anthropic | 92% | $0.230 | 400 |
How is cost calculated?
Cost = (input tokens x prompt price) + (output tokens x completion price). Pricing is fetched from the OpenRouter API at benchmark time. Score per dollar = overall percentage / cost in USD.
Cost = (input tokens x prompt price) + (output tokens x completion price). Pricing is fetched from the OpenRouter API at benchmark time. Score per dollar = overall percentage / cost in USD.