diff --git a/web/test/model-capabilities.test.ts b/web/test/model-capabilities.test.ts
index 3e32cfa70..d320615ca 100644
--- a/web/test/model-capabilities.test.ts
+++ b/web/test/model-capabilities.test.ts
@@ -2,7 +2,7 @@ import assert from "node:assert/strict";
 import test from "node:test";
 
 // Bun 直接执行 TypeScript 测试时需要保留扩展名；生产 tsconfig 不包含 test/。
-import { defaultModelCapabilityConfig, normalizeVideoValue } from "../src/lib/model-capabilities.ts";
+import { DEFAULT_VIDEO_PROMPT_MAX_CHARS, defaultModelCapabilityConfig, normalizeVideoValue } from "../src/lib/model-capabilities.ts";
 
 test("switching to MiniMax H3 replaces an unsupported 720p value with 768P", () => {
     const profile = defaultModelCapabilityConfig("minimax-video", "MiniMax-H3").video!;
@@ -13,3 +13,21 @@ test("switching to MiniMax H3 replaces an unsupported 720p value with 768P", ()
         resolution: "768P",
     });
 });
+
+// 视频提示词由「输入框文本 + 连线内容 + 技能上下文」合成，技能上下文预算为 32000，
+// 合成结果远长于用户手输内容。默认上限过小会把正常可用的画布工作流拦在本地预检。
+// 这里锁定默认值本身，避免被改回偏小值（前端放行/后端拒绝的判定必须同源）。
+test("video prompt default allows a composed canvas prompt", () => {
+    assert.equal(DEFAULT_VIDEO_PROMPT_MAX_CHARS, 8000);
+    for (const protocol of [undefined, "seedance-videos-compatible", "agnes-video", "volcengine-ark-video"]) {
+        const profile = defaultModelCapabilityConfig(protocol, "test-model");
+        assert.equal(profile.video!.references.promptMaxChars, DEFAULT_VIDEO_PROMPT_MAX_CHARS);
+    }
+});
+
+test("raising the video default leaves text and image limits untouched", () => {
+    // 只放宽视频默认值，避免顺带改变其它能力的判定口径。
+    const profile = defaultModelCapabilityConfig("seedance-videos-compatible", "sd-2.5");
+    assert.equal(profile.text!.references.promptMaxChars, 32000);
+    assert.equal(profile.image!.references.promptMaxChars, 32000);
+});