GPT-5.2 benchmark ranking on every AI leaderboard
closed weights · top 10 on 2 of 14 · released Dec 11, 2025
14 of 14 boards
| Rank | Board | Score | |||
|---|---|---|---|---|---|
| 11 |  Search ArenaAgents and tools · as “gpt-5.2-search” | Agents and tools | 1,207 | 33 | |
| 12 |  HLEReasoning · as “gpt-5.2-2025-12-11” | Reasoning | 27.8% | 42 | |
| 14 |  BFCLAgents and tools · as “GPT-5.2-2025-12-11 (FC)” | Agents and tools | 55.9% | 81 | |
| 21 | Vision ArenaVision · as “gpt-5.2-chat-latest-20260210” | Vision | 1,278 | 100 | |
| 22 |  GPQA DiamondReasoning · as “gpt-5.2-2025-12-11_xhigh” | Reasoning | 91.4% | 100 | |
| 24 | Text ArenaChat · as “gpt-5.2-chat-latest-20260210” | Chat | 1,476 | 100 | |
| 24 | MCP AtlasAgents and tools · as “gpt-5.2 (xhigh)” | Agents and tools | 67.6% | 34 | |
| 30 |  ARC-AGI-2Reasoning · as “GPT-5.2 (XHigh)” | Reasoning | 52.9% | 78 | |
| 36 | Epoch ECIReasoning · as “GPT-5.2” | Reasoning | 153.4 | 100 | |
| 36 |  LiveBenchReasoning · as “gpt-5.2-2025-12-11-high” | Reasoning | 74.6% | 59 | |
| 36 | Document ArenaVision · as “gpt-5.2-high” | Vision | 1,405 | 38 | |
| 59 | WebDev ArenaCoding · as “gpt-5.2” | Coding | 1,417 | 100 | |
| 6 |  SWE-bench VerifiedCoding · stale since 2026-02-26 · as “GPT 5.2 (high)” | Coding | 72.8% | 39 | |
| 10 |  MMLU-ProReasoning · stale since 2026-03-11 · as “GPT-5.2” | Reasoning | 87.4% | 100 |
Category ranks and more models
Sources: each leaderboard's own published results page, captured every day at 04:30 UTC. Scores exactly as each board published them; we run no evaluations. Logos via logo.dev; trademarks belong to their owners.