GPT-6 LunavsGPT-5.6 Luna
Across 4 shared benchmarks, GPT-5.6 Luna leads overall: GPT-6 Luna wins 1, GPT-5.6 Luna wins 3, with 0 ties and an average score difference of -2.04.
GPT-6 Luna
OpenAI · 2026-09-22 · Reasoning model
GPT-5.6 Luna
OpenAI · 2026-06-26 · Reasoning model
GPT-6 Luna1 win(25%)(75%)3 winsGPT-5.6 Luna
Benchmark scores
Grouped by capability, sorted by largest gap within each. 4 shared benchmarks.
AI Agent - Tool Usage
GPT-5.6 Luna 1/1| Benchmark | GPT-6 Luna | GPT-5.6 Luna | Diff |
|---|---|---|---|
| Terminal-Bench 4.0 | 1348 / 95Max (With Tools) | 17.2740 / 95Max (With Tools) | -4.27 |
Coding and Software Engineer
GPT-6 Luna 1/1| Benchmark | GPT-6 Luna | GPT-5.6 Luna | Diff |
|---|---|---|---|
| SciCode | 5539 / 134Max (No Tools) | 53.6051 / 134Max (No Tools) | +1.40 |
General Knowledge
GPT-5.6 Luna 1/1| Benchmark | GPT-6 Luna | GPT-5.6 Luna | Diff |
|---|---|---|---|
| ARC-AGI-1 | 86.7067 / 153Max (No Tools) | 8858 / 153Max (No Tools) | -1.30 |
Multimodal Understanding
GPT-5.6 Luna 1/1| Benchmark | GPT-6 Luna | GPT-5.6 Luna | Diff |
|---|---|---|---|
| GDP.pdf | 2040 / 122Max (No Tools) | 2421 / 122Max (No Tools) | -4 |
Specs
| Field | GPT-6 Luna | GPT-5.6 Luna |
|---|---|---|
| Publisher | OpenAI | OpenAI |
| Release date | 2026-09-22 | 2026-06-26 |
| Model type | Reasoning model | Reasoning model |
| Architecture | Dense | Dense |
| Parameters | Not available | Not available |
| Context length | 1.05M | 1.05M |
| Max output | 128K | 128K |
API pricing
Prices use DataLearner records when available; missing fields are not inferred.
| Item | GPT-6 Luna | GPT-5.6 Luna |
|---|---|---|
| Text input | $0.1 / 1M tokens | $0.2 / 1M tokens |
| Text output | $0.5 / 1M tokens | $1.2 / 1M tokens |
| Cache read | $0.01 / 1M tokens | $0.02 / 1M tokens |
| Cache write | $0.125 / 1M tokens | $0.25 / 1M tokens |
Summary
- GPT-6 Lunaleads in:Coding and Software Engineer (1/1)
- GPT-5.6 Lunaleads in:AI Agent - Tool Usage (1/1), General Knowledge (1/1), Multimodal Understanding (1/1)
On average across the 4 shared benchmarks, GPT-5.6 Luna scores 2.04 higher.
Largest single-benchmark gap: Terminal-Bench 4.0 — GPT-6 Luna 13 vs GPT-5.6 Luna 17.27 (-4.27).
Page generated from structured model, pricing and benchmark records. No real-time LLM is used to write the prose.