Automated page. Written and translated automatically by AI and checked against the sources by software; no human editor. SI-IQ is an index built from public benchmark results, not an IQ test. How it works

How smart is GPT-5.2?

125 SI-IQ · 90% interval 122–128
Rank 30 of 80 ranked models · Plausible rank range 15–41

GPT-5.2 is a model from OpenAI, released on Dec 11, 2025. Its SI-IQ is 125 (90% interval 122–128), estimated from 10 benchmark results by 3 independent evaluators. Its strongest area is Math and its weakest is Agents.

Facts

Developer
OpenAI
Released
Dec 11, 2025
Access
API access
Benchmarks used
10 results from 3 evaluators
API price
$1.75 input / $14 output per 1M tokens (developer's price list, Oct 18, 2026)
Context window
272,000 tokens

Ability by area (SI-IQ, 90% interval)

6080100120140160Reasoning126Math126Knowledge124Agents117

Every benchmark result we used

Each score is one value exactly as the source published it: for Epoch AI the best of its run settings, for OpenRouter the median hosting provider (the developer's own hosting when it ran the test). “Implied SI-IQ” is the ability that this single result points to; the overall SI-IQ is their precision-weighted average.

Benchmark Area Score Implied SI-IQ Evaluated by As listed by the source Retrieved Source
Chess Puzzles Reasoning 48.0% 132 Epoch AI gpt-5.2-2025-12-11_xhigh Oct 18, 2026 Epoch AI Benchmarking Hub · data file
Furniture Assembly not in index Reasoning 38.3% — Epoch AI gpt-5.2-2025-12-11_xhigh Oct 18, 2026 Epoch AI Benchmarking Hub · data file
Mystery Game Puzzles Reasoning 23.0% 117 Epoch AI gpt-5.2-2025-12-11_high Oct 18, 2026 Epoch AI Benchmarking Hub · data file
FrontierMath Tier 4 Math 31.7% 126 Epoch AI gpt-5.2-2025-12-11_xhigh Oct 18, 2026 Epoch AI Benchmarking Hub · data file
FrontierMath Tiers 1–3 Math 67.4% 127 Epoch AI gpt-5.2-2025-12-11_xhigh Oct 18, 2026 Epoch AI Benchmarking Hub · data file
OTIS Mock AIME 2024–2025 Math 96.1% 125 Epoch AI gpt-5.2-2025-12-11_high Oct 18, 2026 Epoch AI Benchmarking Hub · data file
SWE-bench Verified (Epoch AI run) Coding 73.8% 119 Epoch AI gpt-5.2-2025-12-11_high Oct 18, 2026 Epoch AI Benchmarking Hub · data file
GPQA Diamond Knowledge 91.4% 127 Epoch AI gpt-5.2-2025-12-11_xhigh Oct 18, 2026 Epoch AI Benchmarking Hub · data file
GPQA Diamond (OpenRouter run) Knowledge 87.9% 125 OpenRouter openai/gpt-5.2-20251211
OpenAI · first-party provider (1 of 1 rows)
Oct 18, 2026 OpenRouter benchmark datasets · data file
SimpleQA Verified Knowledge 37.1% 116 Epoch AI gpt-5.2-2025-12-11_xhigh Oct 18, 2026 Epoch AI Benchmarking Hub · data file
Berkeley Function Calling Leaderboard Agents 55.9% 105 UC Berkeley (Gorilla) GPT-5.2-2025-12-11 (FC) Oct 18, 2026 Berkeley Function Calling Leaderboard · data file
τ²-bench Airline (OpenRouter run) Agents 72.7% 125 OpenRouter openai/gpt-5.2-20251211
OpenAI · first-party provider (1 of 2 rows)
Oct 18, 2026 OpenRouter benchmark datasets · data file

Human preference (Arena)

On the Arena leaderboard, where people vote between two anonymous answers, GPT-5.2 has a rating of 1,435 (1,432–1,439) from 83,855 votes (published Oct 15, 2026). Human preference is a different measure from benchmark ability and is not part of SI-IQ. Arena (Hugging Face)

Models with similar SI-IQ

ModelSI-IQ
Qwen3.7-MaxAlibaba125GPT-5.2 vs Qwen3.7-Max
GLM-5.3Z.ai (Zhipu AI)125GPT-5.2 vs GLM-5.3
Gemini 3.6 FlashGoogle DeepMind125GPT-5.2 vs Gemini 3.6 Flash
Claude Sonnet 5Anthropic126GPT-5.2 vs Claude Sonnet 5
Grok 4.6xAI126GPT-5.2 vs Grok 4.6

Computed with SI-IQ method version 1.0.0. Read the full methodology

Data sources: Arena leaderboard dataset on Hugging Face (CC BY 4.0); Berkeley Function Calling Leaderboard, Gorilla project (Apache-2.0); Epoch AI, ‘Capabilities & benchmarking’ (CC BY 4.0); API list prices via LiteLLM's price table (MIT); each price links to the developer's pricing page; Source: OpenRouter (openrouter.ai), licensed under CC BY 4.0: reuse and republish with attribution to OpenRouter..