VPCT leaderboard
- GPT-5.2 leads
- 84.0%
- models measured
- 15
- lead over #2
- +11.4 pts
- median
- 40.2%
Leader on VPCT · read October 21
On October 21, 2026, 18 of our catalog's models had a VPCT result, and OpenAI GPT-5.2 led with 84.0%, ahead of Google Gemini 3 Flash Preview at 72.6%.
Epoch AI collects its results; its benchmarking hub describes the tasks and links the authors.
Role in the index
VPCT is shown as published and does not feed any of our indexes. Read the methodology
| # | Model | Result | Evidence | Coding index | Output $/Mtok |
|---|---|---|---|---|---|
| 1 | GPT-5.2OpenAI | 84.0% | 55.9 | $14.00 | |
| 2 | Gemini 3 Flash PreviewGoogle | 72.6% | 43.7 | $3.00 | |
| 3 | GPT-5OpenAI | 66.0% | 38.8 | $10.00 | |
| 4 | GPT-5.1OpenAI | 58.7% | 40.4 | $10.00 | |
| 5 | o4 MiniOpenAI | 57.5% | Not benchmarked | $4.40 | |
| 6 | o3OpenAI | 52.0% | Not benchmarked | $8.00 | |
| 7 | Gemini 2.5 Pro Preview 06-05Google | 46.4% | Not benchmarked | $10.00 | |
| 8 | GPT-5 MiniOpenAI | 40.2% | 34.2 | $2.00 | |
| 9 | Claude Opus 4.5Anthropic | 40.0% | 44.0 | $25.00 | |
| 10 | GPT-4o (2024-11-20)OpenAI | 40.0% | Not benchmarked | $10.00 | |
| 11 | Claude Sonnet 4.5Anthropic | 39.8% | 39.5 | $15.00 | |
| 12 | Claude 3.7 SonnetRetiredAnthropic | 39.0% | Not benchmarked | $15.00 | |
| 13 | Claude Opus 4RetiredAnthropic | 38.0% | Not benchmarked | $75.00 | |
| 14 | GPT-5 NanoOpenAI | 37.2% | Not benchmarked | $0.40 | |
| 15 | o1OpenAI | 37.0% | Not benchmarked | $60.00 | |
| 16 | Claude Opus 4.1Anthropic | 35.0% | Not benchmarked | $75.00 | |
| 17 | GPT-4o-mini (2024-07-18)OpenAI | 34.0% | Not benchmarked | $0.60 | |
| 18 | Claude 3.5 SonnetRetiredAnthropic | 33.0% | Not benchmarked | $15.00 |
Only models in our catalog are listed, and each appears once with its best published run. 215 rows were read from the source on October 21, 2026, and rows that match no catalog model are left out.