Model Details
GPT-5.2
Quick answer: GPT-5.2 is a major update in OpenAI's GPT-5 series, succeeding GPT-5.1. It continues the series trajectory of improving agentic coding reliability, reasoning quality, and multi-step task performance while maintaining the $1.25/$10 per million token pricing and 400K+ context window.
What GPT-5.2 Is
GPT-5.2 is OpenAI's next major GPT-5 series release following GPT-5.1. While the .1 update focused on refining the GPT-5 base, the .2 release represents a more significant improvement — likely including an updated knowledge cutoff, improved benchmark scores on coding and reasoning tasks, and refinements to the fast model / thinking model router system introduced in GPT-5.
For teams building on OpenAI's API, GPT-5.2 offers a forward-compatible upgrade path from GPT-5 and GPT-5.1, accessible via dated model version strings for production stability.
Specifications
| Field | Value |
|---|---|
| Organization | OpenAI |
| Context window | 400,000+ tokens |
| License | Proprietary |
| Modality | Multimodal (text and vision) |
Pricing
| Input (per 1M tokens) | Output (per 1M tokens) | |
|---|---|---|
| OpenAI | $1.25 | $10.00 |
Related Models & Benchmarks
- Compare: GPT-5.1 · GPT-5 · Claude Opus 4.6
- Benchmarks: SWE-bench Verified
GPT-5.2 vs Alternatives
| Model | ARC-AGI | ARC-AGI-v2 | BrowseComp | License |
|---|---|---|---|---|
| GPT-5.2 | 86.2% | 52.9% | 65.8% | Proprietary |
| GPT-5.1 | — | — | — | Proprietary |
| Kimi K3 | — | — | 91.2% | Apache 2.0 |
GPT-5.2 BrowseComp (65.8%) trails Kimi K3 (91.2%). For ARC-AGI: GPT-5.2 (86.2%) is competitive. Check for updated comparisons.
Frequently Asked Questions
What is GPT-5.2?
OpenAI's April 2026 model scoring 86.2% ARC-AGI, 65.8% BrowseComp. Proprietary — major GPT-5 series update.Scores from Benchgen evaluations. Last updated 2026-07-24.
Benchmark Leaderboards
This model isn’t on any benchmark leaderboard yet.