GPT-5.6-luna
gpt-5.6-luna
OpenAI · chat · api
Intelligence
Context
1.1M
Max output
128K
Input $/1M
$1.00
Output $/1M
$6.00
Modalities
text
Released
23 Jun 2026
Intelligence Index via Artificial Analysis · 0–100, higher is better
AI summary
● machine-written
GPT-5.6-Luna: OpenAI's low-cost frontier-family tier
GPT-5.6-Luna is the lowest-cost variant in OpenAI's new GPT-5.6 family, positioned for high-volume, latency-sensitive production workloads. It costs $1.00 per 1M input tokens and $6.00 per 1M output tokens, undercutting older budget tiers while maintaining frontier-model capabilities. Luna ships as part of a three-tier release (Sol, Terra, Luna) in limited preview via API access.
What's new
- New $1/$6 per 1M token pricing tier, positioned between GPT-5.4 and GPT-5.4-mini
- Part of GPT-5.6 family limited preview launch (Sol $5/$30, Terra $2.50/$15, Luna $1/$6)
- 1,050,000 token context window with 128,000 token maximum output
- Supports reasoning effort controls and ultra mode for distributed inference
- Available via OpenAI API only; not yet in ChatGPT
Best for
High-volume, cost-sensitive production workloadsLatency-sensitive applications where Sol-grade reasoning is unnecessaryEveryday usage at lower cost than flagship modelsRepeated system prompts and long-context agents with prompt caching