FLUX.2 Pro - Kyma API

Overview

FLUX.2 Pro is BFL’s 32B parameter image model, built for photorealistic generation and unified generation-and-editing workflows. It accepts text prompts and reference images, supporting blends of up to ten source inputs. The model achieves approximately 60% accuracy when rendering text inside generated images. On Kyma, the model runs through an OpenAI-compatible interface using a single API key for all endpoints. Requests receive automatic failover if a serving path degrades, and responses return the exact cost in usage.cost alongside the X-Kyma-Model header. Prompt caching is enabled, billing repeated prompt prefixes at this model’s cached input rate. The model operates at a medium speed tier and does not support structured outputs or reasoning chains. It outputs images only and cannot generate text, code, or audio. Complex multi-reference layouts may require iterative prompting to align precisely with source compositions.

Specs

Field	Value
Model ID	`flux-2-pro`
Best for	Photo realism, multi-reference blend, hero shots, gen+edit unified
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

Photorealistic Product Shots — Generate high-fidelity marketing imagery from text prompts and reference photos.
Multi-Reference Image Blending — Combine up to ten source images into a single cohesive visual composition.
Unified Generation and Editing — Apply targeted edits to existing images within the same generation pipeline.
Text-Integrated Visual Assets — Create images containing rendered text with moderate accuracy for banners or mockups.

Not ideal for

Do not use this model for tasks requiring fast batch processing, structured data extraction, or precise typography beyond basic labels.

Pick something else when

You need faster generation for rapid prototyping: use flux-1.1-ultra or imagen-4-fast.
You require precise typography or layout control: use recraft-v4-pro or ideogram-v3.
You need scalable vector graphics instead of raster images: use recraft-v4-vector or recraft-v4-vector-pro.

Example

from openai import OpenAI

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

response = client.chat.completions.create(
    model="flux-2-pro",
    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

Can I use the same API key for FLUX.2 Pro as I do for text models? Yes, Kyma uses a single API key across all models with an OpenAI-compatible base URL. How does prompt caching affect billing for this model? Repeated prompt prefixes are billed at this model’s cached input rate, reducing costs for iterative editing workflows. Does the model support structured JSON outputs? No, FLUX.2 Pro only outputs images and does not support structured output formatting or reasoning chains.

​Overview

​Specs

​Pricing

​Use this when

​Not ideal for

​Pick something else when

​Example

​Agent query example

​FAQ