diff --git a/docs/protocol.md b/docs/protocol.md
index 4ffe5684..c4255064 100644
--- a/docs/protocol.md
+++ b/docs/protocol.md
@@ -1062,7 +1062,16 @@ thinking level 取值 MINIMAL=1、LOW=2、MEDIUM=3、HIGH=4，由 `reasoning_eff
 
 视频设置来自 Interactions `response_format` 的 `{"type":"video","aspect_ratio","duration","resolution"}` 与 `generation_config.video_config.task`，或 Gemini API `generationConfig.responseFormat` 的 `video.aspectRatio`、`video.duration`、`video.resolution` 与扩展字段 `generationConfig.videoConfig.task`；Interactions `response_modalities` 的 `video` 与 Gemini API `generationConfig.responseModalities` 的 `VIDEO` 等同于不带设置的视频输出格式；`duration` 取 `"4s"` 或秒数并四舍五入为整秒，宽高比与分辨率换成最接近的上游取值，无法解析的取值与未知任务返回 400。
 
-响应为 `[[<EVENT>...], <STATUS>?]`。事件索引 10 的 content delta 中 field 1 为正文、field 5 为视频 `[1,"<BASE64 MP4>"]`、field 6 为思考摘要，分别映射为 text、`video/mp4` 媒体与 reasoning 事件；索引 19 的最终 interaction 状态 `3` 产生 usage（输入、输出、思考、总 token 位于 usage 索引 0、4、8、9）与 `stop` 终态，状态 `4`、`5` 返回错误。事件列表后的 google.rpc 状态按 code 映射 HTTP 状态，额度 code 8 为 429；首个正文事件之前的错误进入换号与冷却，之后的错误以流内 error 结束。
+响应为 `[[<EVENT>...], <STATUS>?]`。事件索引 10 的 content delta 中 field 1 为正文、field 2 至 5 为图片、音频、文档与视频 `[<MIME 枚举>,"<BASE64>"]`、field 6 为思考摘要，分别映射为 text、媒体与 reasoning 事件；索引 19 的最终 interaction 状态 `3` 产生 usage（输入、输出、思考、总 token 位于 usage 索引 0、4、8、9）与 `stop` 终态，状态 `4`、`5` 返回错误。事件列表后的 google.rpc 状态按 code 映射 HTTP 状态，额度 code 8 为 429；首个正文事件之前的错误进入换号与冷却，之后的错误以流内 error 结束。
+
+媒体 MIME 枚举与官网前端相同，枚举为 `null` 或不在表中时取该类的默认 MIME：
+
+| 字段 | 枚举 | 默认 |
+| --- | --- | --- |
+| 图片 field 2 | 1 png、2 jpeg、3 webp、4 heic、5 heif、7 gif、8 bmp、9 tiff | `image/jpeg` |
+| 音频 field 3 | 1 wav、2 mp3、3 aiff、4 aac、5 ogg、6 flac、7 mpeg、8 m4a、9 l16、10 s16le、11 opus、12 alaw、13 mulaw | `audio/wav` |
+| 文档 field 4 | 1 pdf | `application/pdf` |
+| 视频 field 5 | 1 mp4、2 mpeg、3 mpg、4 mov、5 avi、6 x-flv、7 webm、8 wmv、9 3gpp | `video/mp4` |
 
 ### 后台 Interaction 任务
 
diff --git a/internal/aistudio/interaction.go b/internal/aistudio/interaction.go
index 46c51494..4333b460 100644
--- a/internal/aistudio/interaction.go
+++ b/internal/aistudio/interaction.go
@@ -21,11 +21,19 @@ var interactionThinkingLevels = map[int64]int64{4: 1, 1: 2, 2: 3, 3: 4}
 // interactionCancelTimeout 是取消后台任务的期限，包含等待账户请求槽位的时间
 const interactionCancelTimeout = time.Minute
 
-// interactionVideoMIME 表示 Interaction 视频内容的 MIME 枚举
-var interactionVideoMIME = map[int64]string{1: "video/mp4"}
+// interactionMediaMIME 是 content delta 各媒体字段的 MIME 枚举表，与官网前端相同
+var interactionMediaMIME = map[int]interactionMIME{
+	1: {codes: map[int64]string{1: "image/png", 2: "image/jpeg", 3: "image/webp", 4: "image/heic", 5: "image/heif", 7: "image/gif", 8: "image/bmp", 9: "image/tiff"}, unlisted: "image/jpeg"},
+	2: {codes: map[int64]string{1: "audio/wav", 2: "audio/mp3", 3: "audio/aiff", 4: "audio/aac", 5: "audio/ogg", 6: "audio/flac", 7: "audio/mpeg", 8: "audio/m4a", 9: "audio/l16", 10: "audio/s16le", 11: "audio/opus", 12: "audio/alaw", 13: "audio/mulaw"}, unlisted: "audio/wav"},
+	3: {codes: map[int64]string{1: "application/pdf"}, unlisted: "application/pdf"},
+	4: {codes: map[int64]string{1: "video/mp4", 2: "video/mpeg", 3: "video/mpg", 4: "video/mov", 5: "video/avi", 6: "video/x-flv", 7: "video/webm", 8: "video/wmv", 9: "video/3gpp"}, unlisted: "video/mp4"},
+}
 
-// interactionImageMIME 表示 Interaction 图片内容的 MIME 枚举
-var interactionImageMIME = map[int64]string{1: "image/png"}
+// interactionMIME 是一类媒体的 MIME 枚举与枚举缺省或不在表中时的 MIME
+type interactionMIME struct {
+	codes    map[int64]string
+	unlisted string
+}
 
 // interactionStatusHTTP 把流尾 google.rpc 状态码换算为 HTTP 状态
 var interactionStatusHTTP = map[int64]int{
@@ -485,7 +493,7 @@ func decodeInteractionEvent(method string, raw json.RawMessage, emit func(Event)
 	return true, emit(Event{Kind: EventFinish, FinishReason: "stop"})
 }
 
-// decodeInteractionDelta 解码 content delta：field 1 文本、field 2 图片、field 5 视频、field 6 思考摘要、field 7 思考签名、field 10 引用
+// decodeInteractionDelta 解码 content delta：field 1 文本、field 2 图片、field 3 音频、field 4 文档、field 5 视频、field 6 思考摘要、field 7 思考签名、field 10 引用
 func decodeInteractionDelta(method string, raw json.RawMessage, emit func(Event) error) error {
 	delta, err := rawArray(raw, "$delta", raw)
 	if err != nil || len(delta) < 2 {
@@ -512,12 +520,8 @@ func decodeInteractionDelta(method string, raw json.RawMessage, emit func(Event)
 			if err := emit(Event{Kind: EventText, Text: decoded}); err != nil {
 				return err
 			}
-		case 1, 4:
-			mimes := interactionImageMIME
-			if index == 4 {
-				mimes = interactionVideoMIME
-			}
-			media, err := decodeInteractionMedia(method, value, index, mimes, raw)
+		case 1, 2, 3, 4:
+			media, err := decodeInteractionMedia(method, value, index, raw)
 			if err != nil {
 				return err
 			}
@@ -738,20 +742,23 @@ func decodeInteractionCitations(method string, value json.RawMessage, raw json.R
 	return nil
 }
 
-// decodeInteractionMedia 解码图片或视频 content：MIME 枚举与 Base64 数据
-func decodeInteractionMedia(method string, value json.RawMessage, index int, mimes map[int64]string, raw json.RawMessage) (*Blob, error) {
+// decodeInteractionMedia 解码媒体 content：MIME 枚举与 Base64 数据，枚举缺省或不在表中时按官网取该类媒体的默认 MIME
+func decodeInteractionMedia(method string, value json.RawMessage, index int, raw json.RawMessage) (*Blob, error) {
 	path := fmt.Sprintf("$delta[1][%d]", index)
 	media, err := rawArray(value, path, raw)
 	if err != nil || len(media) < 2 {
 		return nil, &ProtocolEvidenceError{Method: method, Path: path, Detail: "媒体 delta 无效", Raw: cloneRaw(raw)}
 	}
-	code, err := rawInt64(media[0], path+"[0]", raw)
-	if err != nil {
-		return nil, withMethod(err, method)
+	var code int64
+	if !isJSONNull(media[0]) {
+		if code, err = rawInt64(media[0], path+"[0]", raw); err != nil {
+			return nil, withMethod(err, method)
+		}
 	}
-	mimeType, ok := mimes[code]
+	kind := interactionMediaMIME[index]
+	mimeType, ok := kind.codes[code]
 	if !ok {
-		return nil, &ProtocolEvidenceError{Method: method, Path: path + "[0]", Detail: fmt.Sprintf("未识别的媒体 MIME 枚举 %d", code), Raw: cloneRaw(raw)}
+		mimeType = kind.unlisted
 	}
 	encoded, err := rawString(media[1], path+"[1]", raw)
 	if err != nil {
diff --git a/internal/aistudio/interaction_media_test.go b/internal/aistudio/interaction_media_test.go
new file mode 100644
index 00000000..76a8b3a6
--- /dev/null
+++ b/internal/aistudio/interaction_media_test.go
@@ -0,0 +1,53 @@
+package aistudio
+
+import (
+	"bytes"
+	"encoding/json"
+	"strings"
+	"testing"
+)
+
+// TestDecodeInteractionMediaMIME 核对各类媒体 delta 按官网枚举表取 MIME，枚举缺省或不在表中时取该类的默认 MIME
+func TestDecodeInteractionMediaMIME(t *testing.T) {
+	data := []byte("\x00\x00\x00\x18ftypmp42\x00\x00\x00\x00mp42isom")
+	for _, tc := range []struct {
+		field int
+		code  any
+		mime  string
+	}{
+		{4, nil, "video/mp4"},
+		{4, 1, "video/mp4"},
+		{4, 7, "video/webm"},
+		{4, 42, "video/mp4"},
+		{1, nil, "image/jpeg"},
+		{1, 1, "image/png"},
+		{1, 3, "image/webp"},
+		{2, nil, "audio/wav"},
+		{2, 11, "audio/opus"},
+		{3, nil, "application/pdf"},
+	} {
+		content := make([]any, tc.field+1)
+		content[tc.field] = []any{tc.code, data}
+		delta := make([]any, 11)
+		delta[10] = []any{0, content}
+		final := make([]any, 20)
+		final[19] = []any{[]any{nil, 3}}
+		wire, err := json.Marshal([]any{[]any{delta, final}, []any{0}})
+		if err != nil {
+			t.Fatal(err)
+		}
+		var media *Blob
+		err = DecodeInteractionStream("CreateInteractionStream", strings.NewReader(string(wire)), func(event Event) error {
+			if event.Kind == EventMedia {
+				media = event.InlineData
+			}
+			return nil
+		})
+		if err != nil {
+			t.Fatalf("field %d code %v: %v", tc.field, tc.code, err)
+		}
+		if media == nil || media.MIME != tc.mime || !bytes.Equal(media.Data, data) {
+			t.Fatalf("field %d code %v: media=%+v", tc.field, tc.code, media)
+		}
+	}
+}