Text to Image API

Seedream 5.0 lite is ByteDance’s unified multimodal image generation model, built with deeper reasoning and online search capabilities to deliver stronger prompt understanding, better visual reasoning, and more accurate, context-aware image creation.

Seedream 4.5 is ByteDance’s upgraded image creation and editing model, designed to deliver higher consistency and fidelity through stronger reference-image preservation, more accurate multi-image editing, and improved typography and dense text rendering for professional visual creative work.

Seedream 4.0 is ByteDance’s new-generation image creation model that unifies image generation and image editing in a single architecture, built to handle complex multimodal tasks such as knowledge-based generation, visual reasoning, and reference-consistent creation while producing high-definition images at up to 4K resolution.

Seedream 3.0 is ByteDance's high-resolution bilingual image generation model, built for Chinese and English prompt understanding with stronger text rendering, better structural accuracy, improved visual aesthetics, and native high-definition image creation.

Nano Banana 2 is Google’s latest image generation and editing model, also called Gemini 3.1 Flash Image, built to combine fast “Flash” speed with stronger visual quality, sharper instruction-following, and more precise edits.