Pricing
What one image costs you on Tokonomix, per size and quality.
Price per image — GPT Image 2.5 Flare
from $0.0071 per image
Billed per image. The price depends on size and quality; your prompt is included.
| Size | Quality | Per image |
|---|---|---|
| Square · 1024×1024 | Low | $0.0088 |
| Square · 1024×1024 | Medium | $0.0198 |
| Square · 1024×1024 | High | $0.0790 |
| Square · 1024×1024 | Extra high | $0.1405 |
| Square · 1024×1024 | Maximum | $0.3161 |
| Landscape · 1536×1024 | Low | $0.0071 |
| Landscape · 1536×1024 | Medium | $0.0154 |
| Landscape · 1536×1024 | High | $0.0617 |
| Landscape · 1536×1024 | Extra high | $0.1107 |
| Landscape · 1536×1024 | Maximum | $0.2470 |
| Portrait · 1024×1536 | Low | $0.0071 |
| Portrait · 1024×1536 | Medium | $0.0154 |
| Portrait · 1024×1536 | High | $0.0617 |
| Portrait · 1024×1536 | Extra high | $0.1107 |
| Portrait · 1024×1536 | Maximum | $0.2470 |
Editing an image: $0.0198 per image
Capabilities
Availability
Availability
No measurements yet
We haven't recorded enough API calls to show availability stats for this model. Data appears once the model starts receiving live traffic.
Tokonomix benchmark verdicts
First-generation multimodal image model with editing and generation
GPT Image 2.5 Flare marks OpenAI's entry into image generation and manipulation capabilities. This baseline model introduces vision understanding, image generation from text prompts, image editing functionality, and transparent background support. As a first-generation release, users should expect this model to establish the foundation for OpenAI's image generation capabilities rather than represent a mature, iterated product. The multimodal approach combines visual input processing with image output creation, enabling workflows that span understanding and generation tasks. Key capabilities include creating images from textual descriptions, modifying existing images based on instructions, and handling transparency for compositing workflows. Without historical performance data or text-based benchmarks to reference, evaluation centers on the breadth of announced capabilities rather than comparative quality metrics. Users exploring this model should focus on testing across the full range of vision, generation, and editing features to understand its practical strengths and limitations in production environments.
Quality
—
Latency p50
—
Test runs
0