Find your model. Know your cost.
Text, code, image, and video APIs. Compare reference list rates with SinoRouter prices in USD.
Planning a larger workload? Talk to our team →Models and API pricing
Crossed-out amounts are reference list rates. Highlighted amounts are SinoRouter prices. Discounts vary by model and configuration.
| Model | Use Case | Billing Unit | List Price / SinoRouter Price |
|---|---|---|---|
| 01Language & Code ModelsText generation, reasoning, agents, and software development | |||
DeepSeek V4 Pro | Complex reasoning, long-context analysis, coding, and enterprise workflows. | 1M tokens | Input |
DeepSeek V4.1 Flash | Fast reasoning, agents, coding, chat, and high-throughput automation. | 1M tokens | Input |
Qwen3.8-Max | High-end text generation, reasoning, coding, and long-context business applications. | 1M tokens | Input |
Qwen3.8-Flash | Fast text generation, assistants, lightweight agents, and high-throughput automation. | 1M tokens | Input |
Qwen3.7-Max | General text generation, reasoning, coding, and multimodal application backends. | 1M tokens | Input |
Qwen3.7-Plus | Long-context text generation and reasoning for document-heavy business workflows. | 1M tokens | Input |
Kimi K3.1 Pro | Text and code generation for business applications. | 1M tokens | Input |
Kimi K3 | Kimi flagship model for long-horizon coding, knowledge work, agents, and 1M-context tasks. | 1M tokens | Input |
Kimi K2.7 Code | Coding-specialized model for long-context programming and agentic development workflows. | 1M tokens | Input |
GLM-5.3 | Complex software engineering, long-running agents, reasoning, and business automation. | 1M tokens | Input |
GLM-5.3-Flash | High-throughput text generation, business automation, and lightweight agent tasks. | 1M tokens | Input |
GLM-5.2 | General reasoning, coding, agents, and long business tasks. | 1M tokens | Input |
MiniMax M3 | Agentic reasoning, tool use, coding, and long-context business workflows. | 1M tokens | Input |
MiniMax M2.7 | Reasoning, coding, and production agents where balanced performance matters. | 1M tokens | Input |
| 02Video ModelsVideo generation for advertising, products, and creative production | |||
Seedance 2.5 | Video generation with resolution and reference-video pricing tiers. | Per second | Without video reference480P With video reference480P |
Seedance 2.0 | Video generation for ads, product content, and creative production. | Per second | Without video reference480P With video reference480P |
Seedance 2.0 Fast | Seedance Fast pricing for 480P and 720P video workflows. | Per second | Without video reference480P With video reference480P |
Seedance 2.0 Mini | Seedance Mini pricing for 480P and 720P video workflows. | Per second | Without video reference480P With video reference480P |
Kling VIDEO 3.0 | Video generation with resolution and native-audio pricing tiers. | Per second | No native audio720P With native audio720P Charged by output duration at the matching resolution and audio rate. |
MiniMax H3 | Multimodal video generation with image, audio, and video references. | Per second | With or without video reference768P First 5 image inputs per task are free; each additional input image costs $0.025, without a discount. |
wan3.0-video | Wan video generation with 480P, 720P, and 1080P pricing tiers. | Per second | With or without video reference480P |
wan3.0-video-prime | Wan Prime video generation with resolution-based pricing. | Per second | With or without video reference480P |
| 03Image ModelsImage generation for ecommerce, marketing, and visual workflows | |||
Seedream 5.5 | Single-image generation with pricing based on total output pixels. | Per image | ≤ 2.61 MP Reference image inputPer input image |
Seedream 5.0 Pro | Single-image generation with pricing based on total output pixels. | Per image | ≤ 2.61 MP |
Seedream 5.0 Flash | A lower-priced Seedream option for image-generation workflows. | Per image | All resolutions |
Kling Image 3.0 | Image generation for product visuals, marketing, and creative workflows. | Per image | 1K / 2K |
Kling Image 3.0 Omni | Image generation for product visuals, marketing, and creative workflows. | Per image | 1K / 2K |
Wan2.7 Image Pro | High-quality image generation for product, marketing, and creative workflows. | Per image | |
Wan2.7 Image | General image generation where cost and speed matter. | Per image | |
Wan2.6 T2I | Text-to-image generation for automated creative and ecommerce content. | Per image | |
For MiniMax H3, Seedance, and Wan, requests without a video reference are billed by output duration. With a video reference, input video duration plus output duration is charged at the matching per-second rate. H3 and Wan use the same rate with or without a video reference. Kling VIDEO 3.0 retains output-duration billing and its existing audio tiers.
Ready to connect? Display names are not API IDs. Retrieve the IDs enabled for your account.
Kimi K3.1 Pro and Seedream 5.5 prices: November 13, 2026. Other video, Seedream and Kling image prices: October 29, 2026. Other reference rates retain their previously published pricing. Display updated November 13, 2026. Provider prices may change without notice. Prices are before taxes, payment processing fees, regional availability limits, and provider-side promotional changes. Time-limited provider discounts are not reflected unless explicitly stated.
Discounts vary by model and configuration. Highlighted media prices are the published SinoRouter rates; no additional discount is applied. Video benchmark and SinoRouter prices are displayed to four decimal places. Model usage rates exclude taxes, payment fees, and separately quoted services. MiniMax M3 pricing uses its standard service tier for requests with up to 512K input tokens.
Disclaimer: model names and trademarks belong to their respective owners. SinoRouter is an independent AI gateway and does not imply official partnership unless separately stated.
Request Enterprise Pricing
Share your use case, expected monthly volume, and preferred models. We will prepare a tailored access and pricing plan.