Currently listed through these providers:
Model details
GPT-5.6 Luna Pro
GPT-5.6 Luna Pro is the lightest of three Pro-tier configurations that OpenAI revealed alongside an upgraded GeneBench-Pro genomics benchmark, sitting below Terra Pro and Sol Pro in capability. The standard Luna tier already improved on that benchmark from 16.5% to 23.6% in the Pro configuration, while Terra Pro reached 28.5% and Sol Pro 31.5%, establishing Luna Pro as the entry point into the Pro lineup. It is framed as the same underlying model as GPT-5.6 Luna, served with reasoning mode set to a higher-quality setting, which lets it handle complex analytical tasks without jumping to the more expensive siblings.
The model is designed for agentic and document-heavy workflows rather than raw scale. It combines tool calling and structured output for orchestrating external systems, attachments and image input for mixed-media grounding, and PDF handling for documents that would otherwise need to be split apart. In practice, hosting telemetry shows Azure's US endpoint sustaining roughly 286 tokens per second with a 7.75s p50 latency and 99.44% uptime over thirty days, while OpenAI's direct endpoint runs faster on latency but trades off on availability, suggesting that deployment choice meaningfully shapes user experience. This makes Luna Pro a practical fit for teams that want Pro-tier reasoning over long inputs at a lower price point than Terra Pro or Sol Pro.
Quick Info
Powered by- Provider
- Venice AI
- Model key
- openai-gpt-56-luna-pro
- Release date
- Jul 9, 2026
- Last updated
- Jul 9, 2026
- Knowledge cutoff
- 2026-02-16
- Input modalities
- Output modalities
- Capabilities
Cost
- Input token cost
- $1.25
- Output token cost
- $7.50
Limits
- Output tokens
- 128,000 tokens
- Context window
- 1,000,000 tokens