GPT-5.6 Luna is the speed- and cost-oriented variant within the GPT-5.6 generation, released alongside the flagship reasoning model Sol and the balanced everyday model Terra. Its position in the family emphasizes tool use and multi-step workflow completion, making it well suited to high-volume production traffic where throughput and price matter more than maximum reasoning depth. Within ChatGPT it has also been adopted as the default experience for free users, with an explicit Think option routing harder questions to higher-reasoning models, which signals confidence in Luna's everyday competence.

Following an 80% price reduction announced at the end of July 2026, Luna is positioned as a frontier-tier option at a fraction of the historical cost of comparable models, with OpenAI reporting Artificial Analysis Intelligence Index scores that place it among the strongest cost-per-task performers in its class. The combination of tool calling, structured output, attachments, and reasoning support lets teams deploy Luna as a default worker for routing, extraction, summarization, and agentic pipelines, reserving heavier models for the queries that actually need them. On Venice's OpenAI-compatible endpoint, this profile translates directly into a low-friction default for builders who want strong general intelligence at sustained volume.