byteplus-image-mcp
Server ini adalah MCP untuk generate dan edit gambar via BytePlus ModelArk/Seedream, terutama untuk logo, mockup UI/UX, dan ilustrasi dokumen.
Generate gambar dari prompt text-to-image untuk logo, mockup, atau ilustrasi dokumen/proposal.
Auto-routing model berdasarkan
purpose: logo/mockup → model kualitas tinggi; document → model hemat.Image-to-image: edit, restyle, atau fusion hingga 14 gambar referensi via path lokal atau URL.
Multi-gambar per request (1–6 sesuai schema; Lite hingga 15 di README) untuk set gambar konsisten.
Pilih resolusi output 1K/1.5K/2K/3K/4K sesuai model.
Seed reproducible untuk hasil konsisten, watermark opsional, dan auto-download hasil ke folder lokal.
Tool
byteplus_list_modelsuntuk melihat model, harga, resolusi, dan panduan pemilihan.Error handling dalam Bahasa Indonesia dengan solusi konkret; API key aman via environment variable.
Allows generating images via BytePlus ModelArk using ByteDance's Seedream models, with automatic model selection for logos, UI/UX mockups, and document illustrations, plus options for resolution, multiple images, and reproducible seeds.
Click on "Deploy Server".
Wait a few minutes for the server to deploy. Once ready, it will show a "Started" state.
In the chat, type
@followed by the MCP server name and your instructions, e.g., "@byteplus-image-mcp Can you generate a logo for my new coffee shop?"
That's it! The server will respond to your query, and you can continue using it as needed.
Here is a step-by-step guide with screenshots.
BytePlus Image MCP
MCP server untuk generate image via BytePlus ModelArk (keluarga model Dola-Seedream 5.0 dari ByteDance) — dirancang untuk kebutuhan logo, mockup UI/UX, dan ilustrasi dokumen/proposal. Mendukung text-to-image dan image-to-image (edit/restyle/fusion dengan gambar referensi).
✨ Fitur
Auto-routing model cerdas — sebut
purposesaja, model dipilih otomatis:logo/mockup→ Dola-Seedream-5.0-pro (teks multibahasa termasuk Bahasa Indonesia, presisi tinggi)document→ Dola-Seedream-5.0-lite (hemat, resolusi hingga 4K)
Image-to-image — berikan gambar referensi (
image_pathspath lokal /image_urls) untuk edit, restyle, atau fusion beberapa gambar (1–14 referensi)Multi-gambar per request via Lite (mode set gambar konsisten, 1–15)
Resolusi per model: Pro 1K/1.5K/2K • Lite 2K/3K/4K — kombinasi tidak valid ditolak dengan pesan solutif
Seed reproducible untuk hasil konsisten
Auto-download hasil ke folder lokal (ekstensi png/jpg terdeteksi otomatis)
Error handling lengkap dengan solusi konkret (Bahasa Indonesia)
Related MCP server: Volcengine Image Generation MCP Server
📊 Model
Model | Model ID | Kegunaan | Resolusi | Harga | Kuota Gratis* |
Dola-Seedream-5.0-pro |
| Logo, mockup, poster/teks di gambar, precise editing & layer control | 1K, 1.5K, 2K | $0.045/gambar | cek konsol |
Dola-Seedream-5.0-lite |
| Ilustrasi dokumen (hemat), batch set gambar konsisten (hingga 15) | 2K, 3K, 4K | $0.035/gambar | 50 gambar |
*Catatan: Pro hanya mendukung 1 gambar per request. Batch multi-gambar hanya via Lite.
Kuota gratis akun baru BytePlus. Daftar di console.byteplus.com. Aktifkan model di ModelArk → Model activation → tab Media.
🚀 Install
Prasyarat
Akun BytePlus dengan API key (ModelArk → API keys)
Aktifkan model Dola-Seedream-5.0 di ModelArk → Model activation → tab Media
Python ≥ 3.10 dan uv (opsional tapi disarankan)
Cara 1: uvx langsung dari GitHub (tanpa install manual)
{
"mcpServers": {
"byteplus-image": {
"command": "uvx",
"args": ["--from", "git+https://github.com/alecslacker/byteplus-image-mcp", "byteplus-image-mcp"],
"env": {
"BYTEPLUS_API_KEY": "isi-api-key-anda"
}
}
}
}Cara 2: pip install dari GitHub
pip install git+https://github.com/alecslacker/byteplus-image-mcpLalu di config MCP client:
{
"mcpServers": {
"byteplus-image": {
"command": "byteplus-image-mcp",
"env": { "BYTEPLUS_API_KEY": "isi-api-key-anda" }
}
}
}Cara 3: jalankan file langsung (lokal)
git clone https://github.com/alecslacker/byteplus-image-mcp
cd byteplus-image-mcp
pip install mcp httpx pydantic
BYTEPLUS_API_KEY=isi-api-key-anda python src/byteplus_image_mcp/server.py⚙️ Environment Variables
Variable | Wajib | Default | Keterangan |
| ✅ | — | API key dari console BytePlus |
| ❌ |
| Folder simpan gambar |
🛠️ Tools
byteplus_generate_image
Parameter | Tipe | Default | Keterangan |
| string | — | Deskripsi gambar (Inggris atau Indonesia — Pro memahami Indonesia native) |
|
| — | Auto-pilih model (logo/mockup→Pro, document→Lite) |
| string | — | Override manual model ID |
|
|
| Resolusi — Pro: 1K/1.5K/2K, Lite: 2K/3K/4K |
| 1–15 |
| Jumlah gambar; >1 hanya untuk Lite (mode set gambar) |
| list string | — | Path gambar referensi lokal (1–14): jpg/png/webp/bmp/tiff/gif, maks 10MB |
| list string | — | URL gambar referensi (1–14), mis. URL hasil generate sebelumnya |
| int | — | Reproducible |
| bool |
| Download otomatis |
| bool |
| Watermark pada gambar |
Contoh pemakaian:
"Buatkan logo Duta Corpora"
→ byteplus_generate_image({
prompt: "minimalist geometric logo, letter D monogram, gold on navy, flat vector",
purpose: "logo"
})
"Ilustrasi cover proposal"
→ byteplus_generate_image({
prompt: "abstract tech illustration, flowing data lines, blue gradient",
purpose: "document"
})
"4 varian ilustrasi konsisten untuk slide deck"
→ byteplus_generate_image({
prompt: "isometric illustration set of Indonesian village scenes ...",
model: "seedream-5-0-260128",
size: "2K",
count: 4
})
"Edit logo yang sudah ada: ganti warna jadi emas"
→ byteplus_generate_image({
prompt: "change the logo color to luxurious gold, keep the shape and layout unchanged",
image_paths: ["C:/uploads/logo-klien.png"], // gambar referensi
purpose: "logo"
})
"Fusion: model dari gambar 1 memegang produk dari gambar 2"
→ byteplus_generate_image({
prompt: "the person in image 1 holding the product from image 2, keep background unchanged",
image_paths: ["C:/uploads/model.jpg", "C:/uploads/produk.jpg"]
})🖼️ Mode Image-to-Image
Berikan gambar referensi, prompt berubah fungsi menjadi instruksi edit:
Upload gambar ke chat MCP client (mis. TRAE) — agent akan mengambil path file-nya
Agent memanggil tool dengan
image_paths(lokal) atauimage_urls(URL, mis. hasil generate sebelumnya yang masih valid 24 jam)Aturan: maks 14 gambar referensi (jpg/png/webp/bmp/tiff/gif, ≤10MB per gambar); Lite: total referensi + hasil ≤ 15; Pro: hasil selalu 1 gambar
Use case: edit logo klien existing, restyle foto produk, mockup dari sketsa, konsistensi karakter antar-aset.
byteplus_list_models
Lihat daftar model + harga + resolusi + panduan pemilihan.
📁 Output
Gambar tersimpan di BYTEPLUS_IMAGE_OUTPUT_DIR dengan nama {timestamp}-{slug-prompt}-{index}.png|.jpg (ekstensi mengikuti format asli dari API). URL dari API valid 24 jam; file lokal permanen.
🔒 Keamanan
API key hanya via environment variable — tidak pernah di-hardcode
Jangan commit
.envatau key ke repositoryMonitor pemakaian: console.byteplus.com → ModelArk → Usage
📄 Lisensi
MIT — lihat LICENSE.
Available Tools
2 toolsbyteplus_generate_imageA
Generate gambar via BytePlus ModelArk (Seedream) untuk logo, mockup UI/UX, dan ilustrasi dokumen.
Tool ini memanggil API BytePlus images/generations dengan model keluarga Seedream. Pemakaian otomatis mengurangi kuota gratis akun (Seedream 4.5: 200 gambar, Seedream 4.0: 200 gambar) sebelum menyentuh saldo berbayar.
Args: params (GenerateImageInput): Parameter tervalidasi: - prompt (str): Deskripsi gambar (bahasa Inggris disarankan), 3-2000 karakter - model (Optional[ImageModel]): Override model secara eksplisit - purpose (Optional[Purpose]): 'logo'|'mockup' -> Seedream 4.5, 'document' -> Seedream 4.0 - size (ImageSize): '1K'|'2K'|'4K' (default 2K) - count (int): 1-6 gambar per request - save_to_disk (bool): Download hasil ke folder output lokal (default true) - watermark (bool): Watermark ByteDance (default false) - seed (Optional[int]): Untuk hasil reproducible
Returns: str: JSON berisi status, model terpakai, URL gambar (valid 24 jam), dan path lokal jika berhasil di-download.
Contoh sukses: { "status": "success", "model": "seedream-4-5-251128", "count": 1, "images": [ {"url": "https://ark-content-generation...", "local_path": "C:/Users/.../logo-duta.jpg", "size_kb": 335} ], "usage_tokens": 16384 }
Kapan dipakai: - "Buatkan logo ..." -> purpose='logo' - "Mockup halaman utama web ..." -> purpose='mockup' - "Ilustrasi untuk proposal ..." -> purpose='document'
Inputs: params (required). Returns: result
byteplus_list_modelsARead-onlyIdempotent
Tampilkan daftar model Seedream yang tersedia beserta panduan pemilihan dan info kuota.
Returns: str: JSON daftar model dengan model_id, use_case, harga per gambar, dan sisa kuota gratis (informasional).
Gunakan tool ini sebelum byteplus_generate_image jika ragu memilih model.
Inputs: no named parameters. Returns: result
Tool Schema Changelog
Recent tool additions, removals, and schema changes observed during successful MCP inspections.
Loading tool schema changelog…
TDQS
Scored across 2 tools
byteplus_generate_image is solely responsible for creating images, while byteplus_list_models handles model discovery and quota information. There is no functional overlap between the two tools.
Both tool names follow the same byteplus_verb_noun pattern: generate_image and list_models. The naming is consistent, predictable, and clearly conveys each tool's action.
With only two tools, the server is slightly under the typical 3-15 range, but the count is reasonable for a narrow image-generation purpose. Each tool earns its place: generation is the core action, and model listing supports informed model selection.
For the server's stated purpose of text-to-image generation, the surface is complete: list_models lets an agent choose the right model, and generate_image produces the image with appropriate parameters. There are no obvious dead ends in the core workflow.
Related MCP Connectors
Generate marketing images, videos and audios for campaigns, product content, and brand assets.
Commercially licensed image generation and editing with IP indemnification.
Create images and videos from prompts, with options for image mixing, reference images, and start/…
Generate images, video, music, voice and 3D through one API. 30 tools, 200+ models.
Related MCP Servers
- AlicenseNot gradedqualityDmaintenanceEnables AI image generation using Volcano Engine's Seedream 4.0 API with text-to-image, image-to-image, multi-image fusion capabilities, built-in prompt templates, and automatic cloud storage integration.19MIT
- AlicenseBqualityDmaintenanceEnables AI-powered text-to-image generation using Volcengine's API with support for multiple image sizes, customizable parameters like guidance scale and seed, and flexible output formats.1MIT
- AlicenseNot gradedqualityDmaintenanceEnables AI-powered image generation using Volcano Engine's SeeDream 4.0 model. Supports custom sizes, reference images, and automatic prompt generation without complex prompting.7 npmMIT
- FlicenseNot gradedqualityCmaintenanceEnables generating AI images using BytePlus Seedream models (4.5 and 5.0 Lite) via natural language, supporting text-to-image, reference-based editing, and sequential multi-panel sets.-