GPT-5.2
by OpenAItext+image+file->text4 endpoints
Neura Intelligence Index
72.4/ 100
Rank
#69—
Confidence
highBenchmarks
13(33% coverage)
Overview
GPT-5.2 is the latest frontier-grade model in the GPT-5 series, offering stronger agentic and long context perfomance compared to GPT-5.1. It uses adaptive reasoning to allocate computation dynamically, responding quickly...
Capabilities
Text generation
Image understanding
Tool use / Function calling
Structured output
Extended reasoning
Prompt caching
Long context
File input
Deterministic seed
Modalities
Input
FileImageText
Output
Text
Technical Specifications
Context Window
400K tokens
Max Output
128K tokens
Tokenizer
GPT
Content Moderation
Enabled
Uptime (24h)
10000.0%
Country
🇺🇸US
Release Date
December 11, 20259mo
Supported Parameters
include_reasoningmax_completion_tokensmax_tokensreasoningreasoning_effortresponse_formatseedstructured_outputstool_choicetools
Benchmark Performance
general
| Benchmark | Score | Source |
|---|---|---|
| MMMLU | 89.6 | verified |
coding
| Benchmark | Score | Source |
|---|---|---|
| SWE-bench Verified | 80 | verified |
math
| Benchmark | Score | Source |
|---|---|---|
| AIME 2025 | 100 | verified |
| FrontierMath | 40.3 | verified |
reasoning
| Benchmark | Score | Source |
|---|---|---|
| GPQA Diamond | 92.4 | verified |
| ARC-AGI v2 | 52.9 | verified |
| Humanity's Last Exam | 34.5 | verified |
multimodal
| Benchmark | Score | Source |
|---|---|---|
| ScreenSpot Pro | 86.3 | verified |
| CharXiv Reasoning | 82.1 | verified |
| MMMU-Pro | 79.5 | verified |
agent
| Benchmark | Score | Source |
|---|---|---|
| BrowseComp | 65.8 | verified |
| MCP Atlas | 60.6 | verified |
| Toolathlon | 46.3 | verified |
Performance Over Time
Pricing
Per 1M tokens
Input$1.75
Output$14.00
Cache read$0.17
Added
December 10, 2025
Last synced 9/19/2026