GPT-5.2
GPT-5.2 is an API-only (closed weights) language model from OpenAI, released December 11, 2025. The context window is 391K tokens. On the Epoch Capabilities Index it scores 154.5, rank 50 of 256 rated models. API price: $1.75 input / $14 output per million tokens.
Check it against your own PC →
Lab
OpenAI
Released
December 11, 2025
Weights
API-only (closed weights)
Parameters
–
Context
391K
Price / M tokens
$1.75 / $14 in / out
Epoch ECI
154.5 #50
Strengths by category
ModelTape's 0–100 fit on Epoch AI benchmark results.
| Category | Score | Based on |
|---|---|---|
| Reasoning | 17.2 | Chess Puzzles, SimpleBench, Mystery Game Puzzles, Humanity's Last Exam |
| Math | 46.1 | ProofBench, OTIS Mock AIME 2024-2025, FrontierMath-Tiers-1-3-v2-Private, FrontierMath-Tier-4-v2-Private |
| Coding | 52.8 | GSO-Bench, MirrorCode, Terminal Bench, SWE-bench Verified |
| Knowledge | 72.5 | GPQA diamond, SimpleQA Verified |
Benchmark results
| Benchmark (Epoch AI) | Score |
|---|---|
| OTIS Mock AIME | 96.1% |
| GPQA Diamond | 92.2% |
| SWE-bench Verified | 73.8% |
| FrontierMath (Tiers 1–3) | 70.9% |
| Terminal-Bench | 64.9% |
| Chess Puzzles | 45.3% |
| SimpleQA Verified | 37.1% |
| SimpleBench | 35.0% |
| MirrorCode | 32.2% |
| FrontierMath-Tier-4-v2-Private | 31.7% |
| Humanity's Last Exam | 27.8% |
| GSO-Bench | 27.5% |
| EBR-bench | 23.0% |
| Mystery Game Puzzles | 15.2% |
| ProofBench | 15.0% |
| Furniture Assembly | 11.9% |
| Remote Labor Index | 2.5% |
Links
Models with a similar score
Gemini 3.7 Flash · Muse Spark 1.3 · GPT-5.4 · Grok 4.6 · Qwen 3.8 Max · Claude Opus 4.7 · GPT-5.2 Pro · GLM-5.3
More from OpenAI
GPT-6.1 Terra · GPT-6.1 Sol · GPT-6 Sol · GPT-6 Luna · GPT-6 Astra · GPT-5.6 Cyber · Daybreak Red · Daybreak Blue