#159 GPT-6 Luna
low- Cost
- $0.002
- Time
- 19.2s
- Tokens
- 2,598 tok
OpenAI
Release: 2026-09-23
Tested on: 2026-09-23 05:09
openai/gpt-6-luna::low
~400B total (~17B active)MoEClosedEstimated
Summary
GPT-6 Luna scores 7.0 on AI BENCHY and ranks #159. It has 9.6 reliability, a 62.1% pass rate, $0.020 total cost, and 21.96s average response time.
What makes GPT-6 Luna unique: It stands out most in Domain specific, where it ranks #2, while Anti-AI Tricks is its weakest area at #14. Its total benchmark cost is unusually low for its score range.
Model facts
Researched on 2026-09-23
- Parameters
- ~400B total (~17B active)
- Architecture
- MoE
- Availability
- Closed
- License
- -
Best estimate from public evidence; the vendor did not disclose every value. Vendor does not disclose parameter counts or architecture. Low-confidence catalog estimate carries forward the prior openai/gpt-5.6-luna family estimate; the linked announcement establishes identity and availability, not these counts.
7.0
Consistency
8.5
9.6
$0.020
Total Output Tokens
22,605
Total Input Tokens
80,159
Input Price
$0.100 / 1M
Output Price
$0.500 / 1M
Flaky tests
4
Flaky tests had mixed outcomes across runs (at least one pass and one fail).
Hamster playing table tennis
Prompt: Create a detailed SVG illustration of a hamster playing table tennis.
Charts
Choose the first model, then click a second model to open a side-by-side page.
Top Models by Score
Score vs Total Cost
Response Time (avg)
Score vs Response Time (avg)
Total Output Tokens
Score vs Total Output Tokens
Quick Compare
Category Breakdown
| Category | Score | Consistency | Tests Correct |
|---|---|---|---|
| Anti-AI Tricks | 6.9 | 7.9 | |
| Coding | 6.0 | 7.2 | |
| Combined | 7.3 | 5.8 | |
| Data parsing and extraction | 10.0 | 10.0 | |
| Domain specific | 5.3 | 10.0 | |
| General Intelligence | 4.4 | 9.9 | |
| Instructions following | 7.1 | 5.8 | |
| Puzzle Solving | 7.7 | 10.0 | |
| Tool Calling | 10.0 | 10.0 | |
| Trivia | 3.0 | 10.0 |