Plans and real price
Cheapest measured$0.032per million tokens, measuredChatGPT Plus · medium confidence · Jul 30, 2026
API list price$4.00 / $20input / output per million tokens
| Plan | Monthly price | Usable tokens / month | Real price / MTok | Confidence | Data date |
|---|
| ChatGPT Plus | $20 | 616M | $0.032 | medium | Jul 30, 2026 |
| ChatGPT Pro 5x | $100 | 3.1B | $0.032 | medium | Jul 30, 2026 |
| Droid Max | $200 | 2.5B | $0.079 | medium | Sep 29, 2026 |
| Command Code GOAT | $10 | 92.1M | $0.109 | medium | Sep 30, 2026 |
| GPT-5.6 Sol API | — | — | $0.588 | high | Sep 5, 2026 |
Data: real-api-pricing · Sources and licenses
Benchmarks
| Configuration | Harness | Score | Cost per task |
|---|
| GPT-5.6 Sol (max) | — | 47.0 | — |
| GPT-5.6 Sol (xhigh) | — | 44.0 | — |
| GPT-5.6 Sol (high) | — | 42.3 | — |
| GPT-5.6 Sol (medium) | — | 39.2 | — |
| GPT-5.6 Sol (low) | — | 33.5 | — |
| GPT-5.6 Sol (non-reasoning) [AA estimate] | — | 28.3 | — |
| Configuration | Harness | Score | Cost per task |
|---|
| Codex - GPT-5.6 Sol (max) | Codex | 37.3 (33.5–41.1) | — |
| Configuration | Harness | Score | Cost per task |
|---|
| Artificial Analysis - GPT-5.6 Sol (max) | Artificial Analysis | 39.9 | — |
| Artificial Analysis - GPT-5.6 Sol (xhigh) | Artificial Analysis | 24.7 | — |
| Artificial Analysis - GPT-5.6 Sol (high) | Artificial Analysis | 20.7 | — |
| Artificial Analysis - GPT-5.6 Sol (medium) | Artificial Analysis | 14.6 | — |
| Artificial Analysis - GPT-5.6 Sol (low) | Artificial Analysis | 1.0 | — |
| Configuration | Harness | Score | Cost per task |
|---|
| gpt-5.6-sol-xhigh (codex-harness) | Codex | 1619.0 (1613.0–1625.0) | — |
| Configuration | Harness | Score | Cost per task |
|---|
| GPT 5.6 Sol (xHigh) | — | 7.0 (5.7–8.4) | — |
| Configuration | Harness | Score | Cost per task |
|---|
| Codex - GPT-5.6 Sol (max) | Codex | 54.6 | $6.35 |
| Configuration | Harness | Score | Cost per task |
|---|
| GPT-5.6 Sol | OpenDesign | 77.6 | $0.544 |
| Configuration | Harness | Score | Cost per task |
|---|
| mini-swe-agent - gpt-5-6-sol (max) | mini-swe-agent | 72.7 (69.8–75.5) | $8.39 |
| mini-swe-agent - gpt-5-6-sol (xhigh) | mini-swe-agent | 70.7 (69.9–71.6) | $4.70 |
| mini-swe-agent - gpt-5-6-sol (high) | mini-swe-agent | 69.4 (68.0–70.8) | $3.47 |
| mini-swe-agent - gpt-5-6-sol (medium) | mini-swe-agent | 61.1 (59.5–62.6) | $1.86 |
| mini-swe-agent - gpt-5-6-sol (low) | mini-swe-agent | 45.4 (43.0–47.7) | $1.07 |
Data: real-api-pricing · Sources and licenses
Codex Radar runs
| Effort | Harness | Radar IQ | Avg cost (USD) | Avg minutes |
|---|
| low | codex | 81.3 | $1.16 | 10.4 |
| medium | codex | 95.5 | $2.26 | 16.0 |
| high | codex | 96.4 | $3.31 | 19.9 |
| xhigh | codex | 98.2 | $4.09 | 24.4 |
| max | codex | 108.5 | $6.34 | 33.9 |
| ultra | codex | 100.5 | $18 | 51.9 |
Data: codexradar · Sources and licenses
