Image
GA
Proprietary
GPT Image 2
Production work that needs text accuracy, instruction following, and editing
Release2026-06
Context32K
Max output—
ParametersUndisclosed
Local / VRAMAPI only
LicenseProprietary
Overview
GPT Image 2 is a model in the Image category from OpenAI. It was released in 2026-06. The vendor is based in US. Best for: Production work that needs text accuracy, instruction following, and editing.
Notes: Thinking mode expands the prompt internally; Instant mode is low latency.
Strengths
- Top Elo (1338) class
- Thinking and Instant modes
- Strong text rendering and image editing
- Transparent background
Limits
- Closed
- Expensive at high resolution
- Parameters not disclosed
Specifications
| Vendor | OpenAI |
|---|---|
| Category | Image |
| Release | 2026-06 |
| Status | GA |
| License | Proprietary |
| Commercial use | Not specified |
| Context window | 32K |
| Input modalities | Text, Image |
| Output modalities | Image |
| Prompt types | Text, Image, Reference |
| Negative prompts | No |
| Max outputs | 1-10 |
| Resolution | 1024x1024, 1536x1024, 1024x1536, up to 2048px |
| Local deployment | No |
| API access | Yes |
| Pricing | $0.006 (low) / $0.053 (medium) / $0.211 (high) per image |
| Confidence | High |
| Vendor country | US |
| Vendor type | Frontier lab |
Pricing & hardware
$0.006 (low) / $0.053 (medium) / $0.211 (high) per image
API-only: local GPU memory is not applicable. API access is available.