Miruro API v3.0

The ultimate, decrypted, and fully reverse-engineered native Python backend for Miruro.

https://github.com/walterwhite-69/Miruro-API

---

⚠️ Cloudflare Notice (Important)

Miruro's pipe endpoint (/api/secure/pipe) is now protected by Cloudflare.

• Do NOT host on Vercel — Vercel uses datacenter IPs that are hard-blocked by Cloudflare's WAF. Requests to the pipe will always return 403.
• Do NOT host on Render free tier, Railway, or similar PaaS — same problem; shared datacenter IPs get blocked.
• ✅ Host on a VPS (DigitalOcean, Hetzner, Linode, Vultr, etc.) with a clean non-datacenter IP, or run locally on your own machine.

The API uses curl_cffi with Chrome TLS fingerprinting and proper same-origin headers to bypass Cloudflare's bot detection. This works from residential and most VPS IPs, but not from known datacenter ranges.

---

What This Does

Miruro's frontend communicates with its backend through a secure/pipe tunnel that base64-encodes, gzip-compresses, and encrypts every request. This project bypasses all of that and gives you simple, direct REST endpoints to:

1. Search & filter anime with full AniList metadata
2. Get complete anime info — characters, staff, relations, recommendations, trailer, stats, and all metadata in one request
3. Browse collections — trending, popular, upcoming, recent, schedule, and spotlight — all paginated
4. List episodes with decoded episode IDs from multiple providers
5. Get M3U8 streaming URLs for any episode
6. Autocomplete search suggestions for dropdown UIs

No headless browsers, no Selenium — just lightweight async HTTP requests.

Setup

git clone https://github.com/walterwhite-69/Miruro-API.git
cd Miruro-API
pip install -r requirements.txt
uvicorn api:app --host 0.0.0.0 --port 8000

Then open http://localhost:8000/ for interactive API docs.

Windows users: If you see a RuntimeWarning about Proactor event loop, that's a harmless curl_cffi warning — the API works fine.

All Endpoints

🔍 Search & Discovery

Endpoint	Description	Params
GET /search?query={name}	Full-text anime search with rich metadata (20+ fields per result)	query (required), page=1, per_page=20
GET /suggestions?query={name}	Lightweight autocomplete for dropdowns — returns id, title, poster, format, status, year. Max 8 results.	query (required)
GET /filter	Advanced browse/filter by any combination of genre, tag, year, season, format, status, sort	All optional — see below

Filter Parameters

Param	Values
genre	Action, Romance, Comedy, Drama, Fantasy, Sci-Fi, etc.
tag	Isekai, Time Skip, Reincarnation, etc.
year	2025, 2024, etc.
season	WINTER · SPRING · SUMMER · FALL
format	TV · MOVIE · OVA · ONA · SPECIAL
status	RELEASING · FINISHED · NOT_YET_RELEASED · CANCELLED
sort	SCORE_DESC · POPULARITY_DESC · TRENDING_DESC · START_DATE_DESC
page / per_page	Pagination (defaults: 1 / 20, max per_page: 50)

---

📊 Collections (All Paginated)

Endpoint	Description
GET /trending	Currently trending anime
GET /popular	Most popular anime of all time
GET /upcoming	Most anticipated upcoming anime
GET /recent	Currently airing / this season's anime
GET /spotlight	Curated "What's Hot" list (trending + popular)
GET /schedule	Airing schedule for the next week

All collection endpoints accept page and per_page query params and return:

{
  "page": 1,
  "perPage": 20,
  "total": 5000,
  "hasNextPage": true,
  "results": [ ... ]
}

Each anime in results includes 20+ fields: title (romaji/english/native), coverImage, bannerImage, format, season, seasonYear, episodes, duration, status, averageScore, meanScore, popularity, favourites, genres, source, countryOfOrigin, studios, nextAiringEpisode, startDate, endDate, and more.

---

📖 Anime Details

Endpoint	Description
GET /info/{anilist_id}	Complete anime page — everything in one request
GET /anime/{id}/characters	Paginated character list with voice actors
GET /anime/{id}/relations	All related media (sequels, prequels, side stories, spin-offs)
GET /anime/{id}/recommendations	Community recommendations sorted by rating

What /info/{id} Returns

Everything you need to build a full anime detail page:

• Core: id, idMal, title (romaji/english/native), description, coverImage, bannerImage
• Metadata: format, season, seasonYear, episodes, duration, status, source, countryOfOrigin
• Scores: averageScore, meanScore, popularity, favourites, trending
• Taxonomy: genres, tags (with rank & spoiler flag), synonyms, hashtag
• People: characters (25, with voice actors), staff (25, with roles)
• Related: relations (sequels/prequels/etc.), recommendations (10, with ratings)
• Media: trailer (YouTube/Dailymotion), streamingEpisodes, externalLinks
• Stats: scoreDistribution, statusDistribution
• Studios: name, isAnimationStudio, siteUrl
• Dates: startDate, endDate, nextAiringEpisode
• Links: siteUrl, externalLinks (MAL, official site, etc.)

---

▶️ Streaming (3-Step Flow)

To get a video stream, follow these 3 steps in order:

Step 1: Get Episodes — GET /episodes/{anilist_id}

Returns all episodes from multiple providers (kiwi, arc, zoro, hop, etc.) organized by audio type.

{
  "mappings": { "anilistId": 178005, "malId": 56885, "kitsuId": "..." },
  "providers": {
    "kiwi": {
      "episodes": {
        "sub": [
          {
            "id": "watch/kiwi/178005/sub/animepahe-1",
            "number": 1,
            "title": "Episode Title",
            "image": "https://...",
            "airDate": "2026-01-04",
            "duration": 1420,
            "description": "...",
            "filler": false
          }
        ],
        "dub": [ ... ]
      }
    },
    "arc": { ... },
    "zoro": { ... }
  }
}

Step 2: Get Sources [SUPER SIMPLE]

Just take the direct id from the Step 1 response and use it as the URL. No manual parameters or complex IDs needed!

Endpoint: GET /{id}
Example: GET /watch/kiwi/178005/sub/animepahe-1

{
  "streams": [
    { "url": "https://.../master.m3u8", "type": "hls", "quality": "1080p" }
  ],
  "subtitles": [
    { "file": "https://...", "label": "English", "kind": "captions" }
  ],
  "intro": { "start": 0, "end": 90 },
  "outro": { "start": 1300, "end": 1420 }
}

Tip

This endpoint automatically handles decryption, provider selection, and category matching. It returns the direct M3U8/HLS streaming URL and intro/outro timestamps.

Fallback / Detailed Option

If you need manual control, you can use the traditional endpoint: `GET /sources?episodeId=...&provider=...&anilistId=...&category=...`

Step 3: Play

Feed streams[0].url into any HLS player (Video.js, hls.js, VLC, mpv). Subtitles are either hard-subbed (baked into the video for kiwi/pahe) or provided in the subtitles array (VTT links for zoro/arc). Use intro/outro timestamps for skip buttons.

How Cloudflare Bypass Works

The pipe endpoint at miruro.tv/api/secure/pipe is behind Cloudflare. The API bypasses it by:

1. Using curl_cffi with chrome110 impersonation — this replicates Chrome's exact TLS fingerprint (JA3/JA4) at the socket level, making the request look indistinguishable from a real browser
2. Sending the correct same-origin headers: sec-fetch-site: same-origin, sec-fetch-mode: cors, sec-fetch-dest: empty, full sec-ch-ua stack, matching Referer and Origin

This approach works from residential IPs and most VPS providers. Cloudflare's hard block only applies to known datacenter/CDN IP ranges (Vercel, Render, AWS Lambda, Cloudflare Workers, etc.).

Disclaimer

This project is for educational purposes and API integrity research only. The author takes absolutely zero responsibility for network usage. Code contains zero skiddable artifacts.

Author: Walter | GitHub: walterwhite-69