GPT 5.6 Terra
GPT 5.6 Terra is developed by OpenAI. MeetNewAI lists 6 benchmark boards and 5 plans for it.
Plans and real price
Cheapest measured$0.0080per million tokens, measuredChatGPT Pro 20x · medium confidence · 2026-06-10~2026-09-06
API list price$2.00 / $12input / output per million tokens
| Plan | Monthly price | Usable tokens / month | Real price / MTok | Confidence | Data date |
|---|
| ChatGPT Pro 20x | $200 | 25B | $0.0080 | medium | 2026-06-10~2026-09-06 |
| ChatGPT Plus | $20 | 1.2B | $0.017 | medium | Jul 30, 2026 |
| ChatGPT Pro 5x | $100 | 6B | $0.017 | medium | Jul 30, 2026 |
| Droid Max | $200 | 5.1B | $0.040 | medium | Sep 29, 2026 |
| GPT-5.6 Terra API | — | — | $0.304 | high | Sep 5, 2026 |
Data: real-api-pricing · Sources and licenses
Benchmarks
| Configuration | Harness | Score | Cost per task |
|---|
| GPT-5.6 Terra (max) | — | 42.1 | — |
| GPT-5.6 Terra (xhigh) | — | 38.0 | — |
| GPT-5.6 Terra (high) | — | 34.2 | — |
| GPT-5.6 Terra (medium) | — | 30.1 | — |
| GPT-5.6 Terra (low) | — | 27.5 | — |
| GPT-5.6 Terra (non-reasoning) | — | 20.8 | — |
| Configuration | Harness | Score | Cost per task |
|---|
| Codex - GPT-5.6 Terra (max) | Codex | 21.5 (18.3–24.8) | — |
| Configuration | Harness | Score | Cost per task |
|---|
| Artificial Analysis - GPT-5.6 Terra (max) | Artificial Analysis | 35.4 | — |
| Artificial Analysis - GPT-5.6 Terra (xhigh) | Artificial Analysis | 10.1 | — |
| Artificial Analysis - GPT-5.6 Terra (high) | Artificial Analysis | 1.5 | — |
| Artificial Analysis - GPT-5.6 Terra (low) | Artificial Analysis | 1.5 | — |
| Artificial Analysis - GPT-5.6 Terra (medium) | Artificial Analysis | 1.0 | — |
| Artificial Analysis - GPT-5.6 Terra (non-reasoning) | Artificial Analysis | 0.5 | — |
| Configuration | Harness | Score | Cost per task |
|---|
| gpt-5.6-terra-xhigh (codex-harness) | Codex | 1519.0 (1512.4–1525.6) | — |
| Configuration | Harness | Score | Cost per task |
|---|
| GPT 5.6 Terra (xHigh) | — | 0.8 (-0.4–2.1) | — |
| Configuration | Harness | Score | Cost per task |
|---|
| mini-swe-agent - gpt-5-6-terra (max) | mini-swe-agent | 69.6 (67.1–72.2) | $4.95 |
| mini-swe-agent - gpt-5-6-terra (xhigh) | mini-swe-agent | 60.2 (58.1–62.3) | $2.13 |
| mini-swe-agent - gpt-5-6-terra (high) | mini-swe-agent | 53.8 (49.4–58.1) | $1.13 |
| mini-swe-agent - gpt-5-6-terra (medium) | mini-swe-agent | 35.1 (31.7–38.5) | $0.583 |
| mini-swe-agent - gpt-5-6-terra (low) | mini-swe-agent | 24.1 (23.3–24.8) | $0.428 |
Data: real-api-pricing · Sources and licenses
Codex Radar runs
| Effort | Harness | Radar IQ | Avg cost (USD) | Avg minutes |
|---|
| low | codex | 50.9 | $0.439 | 8.0 |
| medium | codex | 59.8 | $0.547 | 9.4 |
| high | codex | 77.2 | $0.898 | 13.6 |
| xhigh | codex | 88.0 | $1.53 | 19.3 |
| max | codex | 100.9 | $2.80 | 30.9 |
| ultra | codex | 99.1 | $7.65 | 41.9 |
Community rating: 10.0 / 10 from 1 ratings in the last 24 hours.
Data: codexradar · Sources and licenses
