Model Overview & Capabilities
GPT-6 Luna is OpenAI’s lightweight high-velocity frontier model launched on October 8, 2026, securing Rank #18 on our leaderboard with an empirical composite score of 39.0. Succeeding the previous GPT-5.6 Luna generation, Luna delivers 39.5% on SWE-bench Verified, 38.8% on Terminal-Bench 2.0, 39.5% on Aider Polyglot, and 78.5% on GPQA Diamond.
With ultra-low pricing of $0.10 per million input tokens and $0.50 per million output tokens, GPT-6 Luna yields an attractive blended rate of $0.20 per million tokens under standard 3:1 input:output weighting, delivering an exceptional value score of $0.0051 per index point. It streams tokens at 195 tokens per second with an initial latency of 0.60 seconds, making it one of the most cost-effective commercial models on the market today.
Featuring a 1,100,000-token context window and 16,384 maximum output tokens, GPT-6 Luna is specifically tailored for background codebase indexing, commit message generation, AST linting, documentation scaffolding, automated changelog synthesis, and high-volume utility automation across modern software development workflows. In production developer environments, Luna serves as an always-on assistant capable of handling repetitive engineering tasks without accumulating excessive cloud compute expenses.
