feat: add music cover command and upgrade to music-2.6

- Add `mmx music cover` command for generating cover versions from reference audio (base64 or URL)
- Upgrade music generate model from music-2.5 to music-2.6-free / music-2.6
- Add native `is_instrumental` and `lyrics_optimizer` API fields (replaces 无歌词 workaround)
- Auto-detect key type (sk-cp-* = Token Plan, sk-api-* = PAYG) to select correct model variant
- Add `models.ts` shared utility for model selection logic
- Update MusicRequest type with new fields: is_instrumental, lyrics_optimizer, audio_url, audio_base64, seed, channel
- Fix NodeJS.ErrnoException type in main.ts stdout EPIPE handler
- Update README, README_CN, and skill/SKILL.md

Made-with: Cursor

Author: RyanLee-Dev

README.md

@@ -21,7 +21,7 @@
 - **Image** — Text-to-image with aspect ratio and batch controls
 - **Video** — Async video generation with progress tracking
 - **Speech** — TTS with 30+ voices, speed control, streaming playback
-- **Music** — Text-to-music with optional lyrics
+- **Music** — Text-to-music with lyrics, instrumental mode, auto lyrics, and cover generation from reference audio
 - **Vision** — Image understanding and description
 - **Search** — Web search powered by MiniMax
 - **Dual Region** — Seamless Global (`api.minimax.io`) and CN (`api.minimaxi.com`) support
@@ -99,9 +99,15 @@ mmx speech voices
 ### `mmx music`
 
 ```bash
-mmx music generate --prompt "Upbeat pop" --lyrics "[verse] La da dee, sunny day"
-mmx music generate --prompt "Jazz" --lyrics "La la la" --out song.mp3
+# Generate with lyrics
+mmx music generate --prompt "Upbeat pop" --lyrics "[verse] La da dee, sunny day" --out song.mp3
+# Auto-generate lyrics from prompt
+mmx music generate --prompt "Indie folk, melancholic, rainy night" --lyrics-optimizer --out song.mp3
+# Instrumental (no vocals)
 mmx music generate --prompt "Cinematic orchestral" --instrumental --out bgm.mp3
+# Cover — generate a cover version from a reference audio file
+mmx music cover --prompt "Jazz, piano, warm female vocal" --audio-file original.mp3 --out cover.mp3
+mmx music cover --prompt "Indie folk" --audio https://example.com/song.mp3 --out cover.mp3
 ```
 
 ### `mmx vision`

README_CN.md

@@ -21,7 +21,7 @@
 - **图像生成** — 文生图，支持比例和批量控制
 - **视频生成** — 异步生成，进度追踪
 - **语音合成** — 30+ 音色、语速调节、流式播放
-- **音乐生成** — 文生音乐，支持自定义歌词
+- **音乐生成** — 文生音乐，支持自定义歌词、纯音乐、自动生词，以及基于参考音频的 Cover 生成
 - **图像理解** — 图片描述与识别
 - **网络搜索** — MiniMax 搜索引擎
 - **双区域** — 国际版（`api.minimax.io`）和国内版（`api.minimaxi.com`）自动切换
@@ -99,9 +99,15 @@ mmx speech voices
 ### `mmx music`
 
 ```bash
-mmx music generate --prompt "欢快的流行乐" --lyrics "[主歌] 啦啦啦，阳光照"
-mmx music generate --prompt "爵士风" --lyrics "啦啦啦" --out song.mp3
+# 带歌词生成
+mmx music generate --prompt "欢快的流行乐" --lyrics "[主歌] 啦啦啦，阳光照" --out song.mp3
+# 自动生成歌词
+mmx music generate --prompt "忧郁的独立民谣，雨夜" --lyrics-optimizer --out song.mp3
+# 纯音乐（无人声）
 mmx music generate --prompt "史诗管弦乐" --instrumental --out bgm.mp3
+# Cover — 基于参考音频生成翻唱版本
+mmx music cover --prompt "爵士钢琴，慵懒女声" --audio-file original.mp3 --out cover.mp3
+mmx music cover --prompt "民谣吉他" --audio https://example.com/song.mp3 --out cover.mp3
 ```
 
 ### `mmx vision`

skill/SKILL.md

@@ -192,7 +192,9 @@ echo "Breaking news." | mmx speech synthesize --text-file - --out news.mp3
 
 ### music generate
 
-Generate music. Model: `music-2.5`. Responds well to rich, structured descriptions.
+Generate music. Responds well to rich, structured descriptions.
+
+**Model:** `music-2.6-free` — unlimited for API key users, RPM = 3.
 
 ```bash
 mmx music generate --prompt <text> [--lyrics <text>] [flags]
@@ -201,8 +203,10 @@ mmx music generate --prompt <text> [--lyrics <text>] [flags]
 | Flag | Type | Description |
 |---|---|---|
 | `--prompt <text>` | string | Music style description (can be detailed) |
-| `--lyrics <text>` | string | Song lyrics with structure tags. Use `"\u65e0\u6b4c\u8bcd"` for instrumental. Cannot be used with `--instrumental` |
+| `--lyrics <text>` | string | Song lyrics with structure tags. Required unless `--instrumental` or `--lyrics-optimizer` is used. |
 | `--lyrics-file <path>` | string | Read lyrics from file. Use `-` for stdin |
+| `--lyrics-optimizer` | boolean | Auto-generate lyrics from prompt. Cannot be used with `--lyrics` or `--instrumental`. |
+| `--instrumental` | boolean | Generate instrumental music (no vocals). Cannot be used with `--lyrics`. |
 | `--vocals <text>` | string | Vocal style, e.g. `"warm male baritone"`, `"bright female soprano"`, `"duet with harmonies"` |
 | `--genre <text>` | string | Music genre, e.g. folk, pop, jazz |
 | `--mood <text>` | string | Mood or emotion, e.g. warm, melancholic, uplifting |
@@ -215,7 +219,6 @@ mmx music generate --prompt <text> [--lyrics <text>] [flags]
 | `--structure <text>` | string | Song structure, e.g. `"verse-chorus-verse-bridge-chorus"` |
 | `--references <text>` | string | Reference tracks or artists, e.g. `"similar to Ed Sheeran"` |
 | `--extra <text>` | string | Additional fine-grained requirements |
-| `--instrumental` | boolean | Generate instrumental music (no vocals). Cannot be used with `--lyrics` or `--lyrics-file` |
 | `--aigc-watermark` | boolean | Embed AI-generated content watermark |
 | `--format <fmt>` | string | Audio format (default: `mp3`) |
 | `--sample-rate <hz>` | number | Sample rate (default: 44100) |
@@ -226,19 +229,62 @@ mmx music generate --prompt <text> [--lyrics <text>] [flags]
 At least one of `--prompt` or `--lyrics` is required.
 
 ```bash
-# Simple usage
+# With lyrics
 mmx music generate --prompt "Upbeat pop" --lyrics "La la la..." --out song.mp3 --quiet
 
+# Auto-generate lyrics from prompt
+mmx music generate --prompt "Upbeat pop about summer" --lyrics-optimizer --out summer.mp3 --quiet
+
+# Instrumental
+mmx music generate --prompt "Cinematic orchestral, building tension" --instrumental --out bgm.mp3 --quiet
+
 # Detailed prompt with vocal characteristics
 mmx music generate --prompt "Warm morning folk" \
   --vocals "male and female duet, harmonies in chorus" \
   --instruments "acoustic guitar, piano" \
   --bpm 95 \
   --lyrics-file song.txt \
   --out duet.mp3
+```
+
+---
+
+### music cover
+
+Generate a cover version of a song based on reference audio.
+
+**Model:** `music-cover-free` — unlimited for API key users, RPM = 3.
+
+```bash
+mmx music cover --prompt <text> (--audio <url> | --audio-file <path>) [flags]
+```
+
+| Flag | Type | Description |
+|---|---|---|
+| `--prompt <text>` | string, **required** | Target cover style, e.g. `"Indie folk, acoustic guitar, warm male vocal"` |
+| `--audio <url>` | string | URL of reference audio (mp3, wav, flac, etc. — 6s to 6min, max 50MB) |
+| `--audio-file <path>` | string | Local reference audio file (auto base64-encoded) |
+| `--lyrics <text>` | string | Cover lyrics. If omitted, extracted from reference audio via ASR. |
+| `--lyrics-file <path>` | string | Read lyrics from file. Use `-` for stdin |
+| `--seed <number>` | number | Random seed 0–1000000 for reproducible results |
+| `--format <fmt>` | string | Audio format: `mp3`, `wav`, `pcm` (default: `mp3`) |
+| `--sample-rate <hz>` | number | Sample rate (default: 44100) |
+| `--bitrate <bps>` | number | Bitrate (default: 256000) |
+| `--channel <n>` | number | Channels: `1` (mono) or `2` (stereo, default) |
+| `--out <path>` | string | Save audio to file |
+| `--stream` | boolean | Stream raw audio to stdout |
+
+```bash
+# Cover from URL
+mmx music cover --prompt "Indie folk, acoustic guitar, warm male vocal" \
+  --audio https://filecdn.minimax.chat/public/d20eda57-2e36-45bf-9e12-82d9f2e69a86.mp3 --out cover.mp3 --quiet
+
+# Cover from local file with custom lyrics
+mmx music cover --prompt "Jazz, piano, slow" \
+  --audio-file original.mp3 --lyrics-file lyrics.txt --out jazz_cover.mp3 --quiet
 
-# Instrumental (use --instrumental flag)
-mmx music generate --prompt "Cinematic orchestral, building tension" --instrumental --out bgm.mp3
+# Reproducible result with seed
+mmx music cover --prompt "Pop, upbeat" --audio https://filecdn.minimax.chat/public/d20eda57-2e36-45bf-9e12-82d9f2e69a86.mp3 --seed 42 --out cover.mp3
 ```
 
 ---

src/commands/music/cover.ts

@@ -0,0 +1,120 @@
+import { readFileSync } from 'fs';
+import { defineCommand } from '../../command';
+import { CLIError } from '../../errors/base';
+import { ExitCode } from '../../errors/codes';
+import { request, requestJson } from '../../client/http';
+import { musicEndpoint } from '../../client/endpoints';
+import { formatOutput, detectOutputFormat } from '../../output/formatter';
+import { saveAudioOutput } from '../../output/audio';
+import type { Config } from '../../config/schema';
+import type { GlobalFlags } from '../../types/flags';
+import type { MusicRequest, MusicResponse } from '../../types/api';
+import { musicCoverModel } from './models';
+
+export default defineCommand({
+  name: 'music cover',
+  description: 'Generate a cover version of a song based on reference audio (music-cover-free)',
+  usage: 'mmx music cover --prompt <text> (--audio <url> | --audio-file <path>) [--lyrics <text>] [--out <path>] [flags]',
+  options: [
+    { flag: '--prompt <text>', description: 'Target cover style, e.g. "Indie folk, acoustic guitar, warm male vocal"' },
+    { flag: '--audio <url>', description: 'URL of the reference audio (mp3, wav, flac, etc. — 6s to 6min, max 50MB)' },
+    { flag: '--audio-file <path>', description: 'Local reference audio file (auto base64-encoded)' },
+    { flag: '--lyrics <text>', description: 'Cover lyrics. If omitted, extracted from reference audio via ASR.' },
+    { flag: '--lyrics-file <path>', description: 'Read lyrics from file (use - for stdin)' },
+    { flag: '--seed <number>', description: 'Random seed 0–1000000 for reproducible results', type: 'number' },
+    { flag: '--format <fmt>', description: 'Audio format: mp3, wav, pcm (default: mp3)' },
+    { flag: '--sample-rate <hz>', description: 'Sample rate: 16000, 24000, 32000, 44100 (default: 44100)', type: 'number' },
+    { flag: '--bitrate <bps>', description: 'Bitrate: 32000, 64000, 128000, 256000 (default: 256000)', type: 'number' },
+    { flag: '--channel <n>', description: 'Channels: 1 (mono) or 2 (stereo, default)', type: 'number' },
+    { flag: '--stream', description: 'Stream raw audio to stdout' },
+    { flag: '--out <path>', description: 'Save audio to file' },
+  ],
+  examples: [
+    'mmx music cover --prompt "Indie folk, acoustic guitar, warm male vocal" --audio https://example.com/song.mp3 --out cover.mp3',
+    'mmx music cover --prompt "Jazz, piano, slow" --audio-file original.mp3 --lyrics-file lyrics.txt --out jazz_cover.mp3',
+    'mmx music cover --prompt "Pop, upbeat" --audio https://example.com/ref.mp3 --seed 42 --out reproducible.mp3',
+  ],
+  async run(config: Config, flags: GlobalFlags) {
+    const prompt = flags.prompt as string | undefined;
+    const audioUrl = flags.audio as string | undefined;
+    const audioFile = flags.audioFile as string | undefined;
+
+    if (!prompt) {
+      throw new CLIError('--prompt is required.', ExitCode.USAGE, 'mmx music cover --prompt <text> --audio <url>');
+    }
+
+    if (!audioUrl && !audioFile) {
+      throw new CLIError(
+        'One of --audio <url> or --audio-file <path> is required.',
+        ExitCode.USAGE,
+        'mmx music cover --prompt <text> --audio <url>',
+      );
+    }
+
+    if (audioUrl && audioFile) {
+      throw new CLIError('Use either --audio or --audio-file, not both.', ExitCode.USAGE);
+    }
+
+    let lyrics = flags.lyrics as string | undefined;
+    if (flags.lyricsFile) {
+      const { readTextFromPathOrStdin } = await import('../../utils/fs');
+      lyrics = readTextFromPathOrStdin(flags.lyricsFile as string);
+    }
+
+    const ts = new Date().toISOString().slice(0, 19).replace(/[T:]/g, '-');
+    const ext = (flags.format as string) || 'mp3';
+    const outPath = (flags.out as string | undefined) ?? `cover_${ts}.${ext}`;
+    const format = detectOutputFormat(config.output);
+
+    const model = musicCoverModel(config);
+    const body: MusicRequest = {
+      model,
+      prompt,
+      lyrics,
+      seed: flags.seed as number | undefined,
+      audio_setting: {
+        format: ext,
+        sample_rate: (flags.sampleRate as number) ?? 44100,
+        bitrate: (flags.bitrate as number) ?? 256000,
+        channel: (flags.channel as number) ?? undefined,
+      },
+      output_format: 'hex',
+      stream: flags.stream === true,
+    };
+
+    if (audioUrl) {
+      body.audio_url = audioUrl;
+    } else {
+      body.audio_base64 = readFileSync(audioFile!).toString('base64');
+    }
+
+    if (config.dryRun) {
+      console.log(formatOutput({ request: body }, format));
+      return;
+    }
+
+    const url = musicEndpoint(config.baseUrl);
+
+    if (flags.stream) {
+      const res = await request(config, { url, method: 'POST', body, stream: true });
+      const reader = res.body?.getReader();
+      if (!reader) throw new CLIError('No response body', ExitCode.GENERAL);
+      while (true) {
+        const { done, value } = await reader.read();
+        if (done) break;
+        process.stdout.write(value);
+      }
+      reader.releaseLock();
+      return;
+    }
+
+    const response = await requestJson<MusicResponse>(config, {
+      url,
+      method: 'POST',
+      body,
+    });
+
+    if (!config.quiet) process.stderr.write(`[Model: ${model}]\n`);
+    saveAudioOutput(response, outPath, format, config.quiet);
+  },
+});