GPT 5.6 Luna
GPT 5.6 Luna is developed by OpenAI. MeetNewAI lists 7 benchmark boards and 7 plans for it.
Plans and real price
Cheapest measured$0.0014per million tokens, measuredChatGPT Pro 20x · medium confidence · Sep 8, 2026
API list price$0.200 / $1.20input / output per million tokens
| Plan | Monthly price | Usable tokens / month | Real price / MTok | Confidence | Data date |
|---|
| ChatGPT Pro 20x | $200 | 141.2B | $0.0014 | medium | Sep 8, 2026 |
| ChatGPT Pro 5x | $100 | 35.3B | $0.0028 | medium | Sep 8, 2026 |
| ChatGPT Plus | $20 | 7.1B | $0.0028 | high | Sep 8, 2026 |
| Droid Max | $200 | 50.5B | $0.0040 | medium | Sep 29, 2026 |
| Command Code GOAT | $10 | 657.9M | $0.015 | medium | Sep 30, 2026 |
| OpenCode Go | $10 | 493.4M | $0.020 | medium | Sep 30, 2026 |
| GPT-5.6 Luna API | — | — | $0.030 | high | Sep 5, 2026 |
Data: real-api-pricing · Sources and licenses
Benchmarks
| Configuration | Harness | Score | Cost per task |
|---|
| GPT-5.6 Luna (max) | — | 37.3 | — |
| GPT-5.6 Luna (xhigh) | — | 34.6 | — |
| GPT-5.6 Luna (high) | — | 32.1 | — |
| GPT-5.6 Luna (medium) | — | 25.0 | — |
| GPT-5.6 Luna (low) | — | 21.0 | — |
| GPT-5.6 Luna (non-reasoning) | — | 15.5 | — |
| Configuration | Harness | Score | Cost per task |
|---|
| Codex - GPT-5.6 Luna (max) | Codex | 17.3 (14.4–20.1) | — |
| Configuration | Harness | Score | Cost per task |
|---|
| Artificial Analysis - GPT-5.6 Luna (max) | Artificial Analysis | 11.6 | — |
| Artificial Analysis - GPT-5.6 Luna (xhigh) | Artificial Analysis | 3.5 | — |
| Artificial Analysis - GPT-5.6 Luna (high) | Artificial Analysis | 2.5 | — |
| Artificial Analysis - GPT-5.6 Luna (non-reasoning) | Artificial Analysis | 1.0 | — |
| Artificial Analysis - GPT-5.6 Luna (medium) | Artificial Analysis | 0.5 | — |
| Artificial Analysis - GPT-5.6 Luna (low) | Artificial Analysis | 0.0 | — |
| Configuration | Harness | Score | Cost per task |
|---|
| gpt-5.6-luna-xhigh (codex-harness) | Codex | 1518.0 (1511.6–1524.4) | — |
| Configuration | Harness | Score | Cost per task |
|---|
| GPT 5.6 Luna (xHigh) | — | -1.1 (-1.8–-0.4) | — |
| Configuration | Harness | Score | Cost per task |
|---|
| Codex - GPT-5.6 Luna (max) | Codex | 43.2 | $0.438 |
| Configuration | Harness | Score | Cost per task |
|---|
| mini-swe-agent - gpt-5-6-luna (max) | mini-swe-agent | 67.2 (63.2–71.2) | $3.03 |
| mini-swe-agent - gpt-5-6-luna (xhigh) | mini-swe-agent | 56.9 (54.7–59.0) | $1.54 |
| mini-swe-agent - gpt-5-6-luna (high) | mini-swe-agent | 44.2 (41.3–47.2) | $0.778 |
| mini-swe-agent - gpt-5-6-luna (medium) | mini-swe-agent | 11.3 (10.5–12.1) | $0.216 |
| mini-swe-agent - gpt-5-6-luna (low) | mini-swe-agent | 1.5 (0.7–2.4) | $0.072 |
Data: real-api-pricing · Sources and licenses
Codex Radar runs
| Effort | Harness | Radar IQ | Avg cost (USD) | Avg minutes |
|---|
| low | codex | 7.1 | $0.026 | 7.1 |
| medium | codex | 32.6 | $0.066 | 10.7 |
| high | codex | 76.3 | $0.162 | 17.3 |
| xhigh | codex | 86.2 | $0.256 | 24.6 |
| max | codex | 102.2 | $0.426 | 37.6 |
Community rating: 8.0 / 10 from 1 ratings in the last 24 hours.
Data: codexradar · Sources and licenses
