invokeable()
← directory / models / gpt-5.6-luna

GPT-5.6 Luna

model gpt-5.6-luna

Draft entry — numbers not yet verified against primary sources.

1.1M
context
128K
max output
—
input
—
output
2026-02
knowledge cutoff
1
providers

01Providers

ProviderProvider model IDIn $/MOut $/MContextStatus
OpenAI API gpt-5.6-luna — — 1.1M live

02Invoke

provider
language

  

  



03Benchmarks · vendor-reported unless verified
SWE-bench Pro resolve rate62.7%




Terminal-Bench (success rate) success rate84.7




DeepSWE resolve rate67.2%




GPQA Diamond (accuracy) accuracy92.3




FrontierMath (accuracy) accuracy · Tier 1-378.6




BrowseComp (accuracy) accuracy83.3




04Related
GPT-5.6 Sol2026-07GPT-5.6 Terra2026-07GPT-5.5 Instant2026-05GPT-5.52026-04GPT-5.5 Pro2026-04GPT-5.4 mini2026-03


05Compare
vs Qwen3.8 27Bhead-to-headvs Gemini 3.7 Flashhead-to-headvs Mistral Medium 3.5head-to-headvs Claude Opus 5head-to-headvs Nemotron 3.5 Lightning 30B A3Bhead-to-headvs GLM-5.3head-to-headvs North Mini Codehead-to-headvs GPT-5.6 Solhead-to-head


06Sources · where these numbers come from
  1. https://github.com/sst/models.dev
  2. https://platform.openai.com/docs/models
  3. https://openai.com/news/
  4. https://platform.openai.com/docs