AI Benchmark Model
GPT-6 Luna (high)
Benchmark scores, pricing, speed, and model comparisons for GPT-6 Luna (high).
Key scores
The GPT-6 Luna (high) model, developed by OpenAI and released on September 22, 2026, is featured in this benchmark evaluation. The model achieves an intelligence index score of 32.1, while specific metrics for coding and math remain unknown. In terms of performance, it demonstrates an output speed of 131.927 tokens per second with a time to first token of 10.661 seconds. Benchmark results for the model include an HLE score of 0.329, a SciCode score of 0.503, and an LCR score of 0.793333333333333. Pricing for the GPT-6 Luna (high) is structured at $0.10 per 1 million input tokens and $0.50 per 1 million output tokens, resulting in a blended rate of $0.20 per 1 million tokens.
Benchmark results
3 tracked benchmark rows for GPT-6 Luna (high).
LCR Other
79.3% SciCode Other
50.3% HLE Other
32.9%