GPT Image 2 - Kyma API

Overview

Released in April 2026, this model accepts text and image inputs to generate images at 1024×1024, 1024×1536, 1536×1024, or 2048×2048 resolutions. It features reasoning-augmented composition and supports low, medium, and high quality tiers, with medium as the default. Kyma routes requests through an OpenAI-compatible endpoint using a single API key. The platform provides automatic failover if a serving path degrades, and every response returns the exact cost in the usage.cost field alongside an X-Kyma-Model header confirming the executed model. Prompt caching is supported, billing repeated prefixes at this model’s cached input rate. The model does not support structured outputs or reasoning capabilities. Generation speed is rated medium, and it operates strictly as an image-to-image or text-to-image generator without native text output capabilities.

Specs

Field	Value
Model ID	`gpt-image-2`
Best for	Text-in-image, photoreal, multilingual typography
Context window	0
Input modalities	Text, Image
Output modalities	Image
Tool calling	Yes
Structured outputs	Yes
Prompt caching	Yes
Speed	Medium
Cost band	Balanced
Release stage	Stable

Pricing

	Per 1M tokens
Input	$0.20
Output	$0.60

Use this when

Multilingual Typography Rendering — Generates accurate text overlays in multiple languages directly within generated images.
Photorealistic Scene Composition — Produces high-fidelity photographs and realistic environments from descriptive prompts.
Logo and Brand Asset Creation — Renders precise, logo-grade lettering and graphic elements suitable for branding workflows.
Reference Image Editing — Modifies existing images using text instructions while preserving core visual structure.

Not ideal for

Do not use this model if you require fast, low-latency generation, structured JSON outputs, or reasoning capabilities.

Pick something else when

You need faster generation times: use imagen-4-fast.
You require scalable vector graphics: use recraft-v4-vector or recraft-v4-vector-pro.
You prioritize ultra-high resolution outputs: use flux-2-pro or imagen-4-ultra.

Example

from openai import OpenAI

client = OpenAI(base_url="https://kymaapi.com/v1", api_key="ky-...")

response = client.chat.completions.create(
    model="gpt-image-2",
    messages=[{"role": "user", "content": "..."}],
)

Agent query example

Ask the API which models fit, instead of hardcoding an id:

curl "https://kymaapi.com/v1/models?recommended_for=generate-images&tools=true&quality_tier=frontier-open"

FAQ

Does Kyma support prompt caching for this model? Yes, repeated prompt prefixes are billed at this model’s cached input rate. What happens if the primary serving path fails? Kyma automatically reroutes your request to a healthy serving path without requiring manual intervention or retry logic. Can I get structured outputs like JSON from this model? No, this model does not support structured outputs; it returns generated images only.

​Overview

​Specs

​Pricing

​Use this when

​Not ideal for

​Pick something else when

​Example

​Agent query example

​FAQ