From 65e943db04fa97871b79d1a068c7afa560e5e688 Mon Sep 17 00:00:00 2001 From: Karen Sheng Date: Sat, 18 Apr 2026 01:14:43 -0700 Subject: [PATCH] Remove broken wan2.7-i2v row; point video users to wan2.6-t2v Live testing confirmed wan2.7-i2v returns 'model not supported' from /apis/v1/services/aigc/video-generation/video-synthesis. The model listed in the catalog at $1.836/video is not actually routed; advertising it leads developers to hit HTTP 500. The model that DOES work today is wan2.6-t2v. Adding a paragraph below the media table directing video-gen users to it via the Video API page, and to email developer@aisa.one for the current per-second rate (the spreadsheet we derived other pricing from did not include wan2.6-t2v). Once the product team confirms the current rate, a follow-up PR can add wan2.6-t2v as a proper row. Co-Authored-By: Claude Opus 4.6 (1M context) --- guides/pricing/ai-model-pricing-llm-inference.mdx | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git a/guides/pricing/ai-model-pricing-llm-inference.mdx b/guides/pricing/ai-model-pricing-llm-inference.mdx index 9041bd4..91c9252 100644 --- a/guides/pricing/ai-model-pricing-llm-inference.mdx +++ b/guides/pricing/ai-model-pricing-llm-inference.mdx @@ -196,7 +196,8 @@ Some models generate media rather than tokens. These models are billed per asset | seedream-4-5-251128 | ByteDance | $0.040 per image | | wan2.7-image | Qwen | $0.030 per image | | wan2.7-image-pro | Qwen | $0.075 per image | -| wan2.7-i2v | Qwen | $1.836 per video (i2v) | + +For **video generation**, the currently-supported model is `wan2.6-t2v` via the async [Video API](/api-reference/video/post_services-aigc-video-generation-video-synthesis). Contact [developer@aisa.one](mailto:developer@aisa.one) for the current per-second rate. ## **Important Notes**