Models

Review supported models, use cases, billing methods, and reference prices. The final estimate is shown before each generation.

Showing 14 models
Text to image

GPT Image 2 · Eaheng

¥0.04 / run

Detailed, polished image generation through BananaRouter.

Image generationPer run
Text to image

GPT Image 2 · Tuzi

¥0.50 / run

Detailed, polished image generation through BananaRouter.

Image generationPer run
Text to image

Z-Image-Turbo

¥0.01 / run

Fast image drafts, product concepts, and style exploration.

Image generationPer run
Text to image

Gemini 3.1 Flash Image

¥4.02 / run

Tuzi original-price Gemini Flash image channel with a one-cent service margin.

Image generationPer run
Text to image

Gemini 3 Pro Image · 4K

¥0.51 / run

Tuzi listed-price Gemini Pro 4K image and reference-image channel with a one-cent service margin.

Image generationPer run
Text to image

Qwen-Image-2512

¥0.01 / run

Chinese prompts, posters, ecommerce images, and campaign visuals.

Image generationPer run
Image understanding

Qwen Reverse Prompt

¥0.01 / run

Reverse prompt extraction model.

Image understandingPer run
Text

Qwen 3.5 Text

¥0.01 / run

Primary text generation model.

TextPer run
Text

GPT 5.6 Luna

¥0.02 / run

Metered Tuzi original-price Luna channel. The displayed amount is a reservation cap; settlement uses reported token usage.

TextPer run
Text

GPT 5.6 Sol

¥0.08 / run

Metered Tuzi original-price text channel. The displayed amount is the reservation cap.

TextPer run
Text

GPT 5.6 Terra

¥0.04 / run

Metered Tuzi original-price Terra channel. The displayed amount is a reservation cap; settlement uses reported token usage.

TextPer run
Text

DeepSeek V4 Flash · 官方

¥0.08 / run

DeepSeek official usage-based channel. The displayed amount is a CNY reservation cap; settlement uses reported cache-hit, cache-miss, and output tokens.

TextPer run
Text

DeepSeek V4 Pro · 官方

¥0.21 / run

DeepSeek official usage-based channel. The displayed amount is a CNY reservation cap; settlement uses reported cache-hit, cache-miss, and output tokens.

TextPer run
Text

MiniMax H3

¥2.01 / run

Canonical MiniMax H3 multimodal video generation: 768P CNY 0.50/second or 2K CNY 0.80/second, plus documented reference-media input charges.

TextPer run