#93 GPT-5.6 Luna
medium- Cost
- $0.014
- Time
- 14.6s
- Tokens
- 2,362 tok
OpenAI
Release: 2026-07-09
Tested on: 2026-08-14 02:20
openai/gpt-5.6-luna::medium
~400B total (~17B active)MoEClosedEstimated
Summary
GPT-5.6 Luna scores 7.4 on AI BENCHY and ranks #93. It has 10.0 reliability, a 62.1% pass rate, $0.072 total cost, and 7.43s average response time.
What makes GPT-5.6 Luna unique: It stands out most in Combined, where it ranks #1, while Coding is its weakest area at #17. Its total benchmark cost is unusually low for its score range.
Model facts
Researched on 2026-08-12
- 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 the architecture or count.
7.4
Consistency
9.2
10.0
Total Output Tokens
44,525
Total Input Tokens
89,685
Input Price
$0.200 / 1M
Output Price
$1.200 / 1M
Flaky tests
2
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.
Run history
| Tested on | Score | Reliability | Tests Correct | Total Cost | Compare |
|---|---|---|---|---|---|
| 2026-08-14 02:20 Re-test | 7.4 | 10.0 | $0.036 ↓ | Current run | |
| 2026-07-16 22:37 New test added | 7.6 | 10.0 | $0.352 | Compare | |
| 2026-07-09 21:16 Initial run | 7.6 | 10.0 | $0.258 | Compare |
This run used a different benchmark suite. Keep suite changes in mind when reading historical movement.
Run comparison
| Run | Benchmark coverage | Score | Consistency | Reliability | Tests Correct | Flaky tests | Total Output Tokens | Total Input Tokens | Total Cost | Response Time (avg) |
|---|---|---|---|---|---|---|---|---|---|---|
| 2026-08-14 02:20 · Current run | 66/66 attempts | 7.4 | 9.2 | 10.0 | 13/22 | 2 | 44,525 | 89,685 | $0.072 | 7.43s |
| 2026-07-09 21:16 · Initial run | 63/63 attempts | 7.6 | 9.6 | 10.0 | 13/21 | 1 | 36,888 | 35,699 | $0.258 | 6.70s |
| Difference | — | -0.1 | -0.4 | 0.0 | 0 | +1 | +7637 | +53986 | -$0.186 | +737ms |
Benchmark coverage differs: 66/66 attempts (Target: 3 repeats per test) versus 63/63 attempts (Target: 3 repeats per test). Totals and repeat-sensitive metrics are not directly comparable.
These two runs used different benchmark suites, so the deltas reflect both model changes and suite changes.
Price History
Historical pricing data for this model from OpenRouter.
| Date | Input Price | Output Price |
|---|---|---|
| 2026-07-09 21:18 | $1.000 / 1M | $6.000 / 1M |
| 2026-07-28 08:40 | $0.500 / 1M | $3.000 / 1M |
| 2026-07-31 10:14 | $0.100 / 1M | $0.601 / 1M |
| 2026-08-14 12:51 | $0.100 / 1M | $0.600 / 1M |
| 2026-08-20 12:51 | $0.200 / 1M | $1.200 / 1M |
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
GPT-5.6 LunamediumvsKimi K2.7 CodemediumGPT-5.6 LunamediumvsLongCat 2.0mediumGPT-5.6 LunamediumvsQwen3.7 MaxnoneGPT-5.6 LunamediumvsQwen3.8 2.4T A95BhighGPT-5.6 LunamediumvsGemini 3.1 Flash Lite PreviewmediumGPT-5.6 LunamediumvsGrok 4.6lowGPT-5.6 LunamediumvsGrok Build 0.1mediumGPT-5.6 LunamediumvsGemini 3 Flash PreviewlowGPT-5.6 LunamediumvsGPT-5.4 MinimediumGPT-5.6 LunamediumvsClaude Sonnet 4.6none
Category Breakdown
| Category | Score | Consistency | Tests Correct |
|---|---|---|---|
| Anti-AI Tricks | 8.3 | 10.0 | |
| Coding | 5.4 | 7.2 | |
| Combined | 10.0 | 10.0 | |
| Data parsing and extraction | 10.0 | 10.0 | |
| Domain specific | 2.9 | 7.2 | |
| General Intelligence | 5.1 | 10.0 | |
| Instructions following | 9.9 | 10.0 | |
| Puzzle Solving | 7.8 | 10.0 | |
| Tool Calling | 10.0 | 10.0 | |
| Trivia | 3.0 | 10.0 |