API Endpoints: https://api.ryzenths.dpdns.org/api/v1/chat/completions

Supports multiple models:
- gpt-5
- gpt-5-mini
- claude-sonnet-4
- gemini-2.0-flash
- grok-4
- deepseek-ai/deepseek-v3.1
- ryzenth-web-compatibility
- moonshotai/kimi-k2-instruct
- meta-llama/llama-4-scout-17b-16e-instruct
- openai/gpt-oss-20b
...and many more.

Authorization Required?

Starting from October 1, all requests will require authorization.

Please don’t DM me 100 times about this too many people are already following up with me.

API Updated Now Live

Endpoint: https://api.ryzenths.dpdns.org/api/v1/images/generations

Supported models:
- Ryzenth-gen-ultra
- Ryzenth-gen-pro
- Ryzenth-gen-creative
- Ryzenth-gen-standard
- Ryzenth-gen-turbo
- Ryzenth-gen-realistic
- Ryzenth-gen-anime
- Ryzenth-gen-3d
- Ryzenth-gen-abstract
- Ryzenth-gen-lite
- Ryzenth-gen-mobile

Note: We do not offer a $5/month plan. The server may return errors during peak load, please be patient and retry if requests fail.

Ghibli API: Access the dedicated Ghibli editor at the following V1 endpoint: https://api.ryzenths.dpdns.org/api/v1/ghibli/edit.

Media Generation (V3): You can test our image and video generation features without an API key using the V3 endpoint: https://api.ryzenths.dpdns.org/api/v3/videos/snaken/generations.

Chat Completions:
Use V1 (https://api.ryzenths.dpdns.org/api/v1/chat/completions) for a wide variety of models, including Gemini, Claude, GPT, and Grok. among others.

Use V2 (https://api.ryzenths.dpdns.org/api/v2/chat/completions) for our latest chat completion features.

Available Models: Our platform hosts powerful models such as ryzenth-neos-v1, ryzenth-quantum-ultra, ryzenth-synapse-max, and ryzenth-nexus-prime, among others.

Important Note:
Please be advised that these services are not free. Access requires authorization. You can obtain a key by contacting our support or by building your own API integration. Free testing is only available for the specified V3 media generation endpoint.

Now Imagen 4 is faster

require "bundler/setup"
require "anthropic"

anthropic = Anthropic::Client.new(
  api_key: ENV["RYZENTH_API_KEY"],
  base_url: "https://api.ryzenths.dpdns.org/api/v1/"
)

message = anthropic.messages.create(
  max_tokens: 1024,
  messages: [{role: "user", content: "Hello, Claude"}],
  model: "claude-sonnet-4"
)

puts(message.content)

Base URL: All requests must be directed to https://api.ryzenths.dpdns.org/api/v1/.

API Key: You must use your purchased RYZENTH_API_KEY for authorization.

Multi-SDK Support: This pattern works with multiple libraries:

- Python & TypeScript: Use the openai library with a custom base_url.

- Go: Use the go-openai library and set the BaseURL field in the client config.

- Gemini: The Google Generative AI SDK can also be configured similarly to use our gateway for Gemini models.

New Features:

New API Endpoint: The primary Chat Completions API is now available at:
https://api.ryzenths.dpdns.org/v2.1/chat/completions

Extended Free Model Access: The following models are now supported in the free tier:

ryzenth-vision-banana
neura-sight-ultra
omni-visio-pro
cortex-view-7b
byte-forge-32b
nexus-chat-ultra
dolphin-mistral-24b-venice-edition
...and several others.

Action Required: Update your integration to use the /v2.1/chat/completions endpoint for compatibility with the new model inventory.

2756 model gila sih mana mau dicoba semua wkwk ☠️

Bjorka: Salah tangkap lagi

WKWKWK GAGUNA BANGET API 5 endpoints (channel Premium) tapi kayak sampah,

API key malah disuruh bayar Miskin amat lu 🤣

Code snippets:

# Gemini Nano banana Free
# Code by Randy

from Ryzenth.helper import to_buffer
import os, requests, re

url = "https://api.ryzenths.dpdns.org/api/v2.1/chat/completions"

headers = {
    "content-type": "application/json"
}
payload = {
    "model": "ryzenth-vision-banana",
    "messages": [{"role": "user", "content": "make a image dog run funny"}]
}

resp = requests.post(url, headers=headers, json=payload)

resp_json = resp.json()
text_content = resp_json["choices"][0]["message"]["content"]
url_field = resp_json["choices"][0]["message"]["images"][0]["image_url"]["url"]
match = re.search(r"^data:image\/\w+;base64,(.+)", url_field)

if match:
    base64_str = match.group(1)
    img = to_buffer(base64_str, return_default_base64=True)
    print("Saved", img)
else:
    print(text_content)

Note:
GOBLOK LU! Channel premium private tapi ribetnya minta ampun, Mending RENDYPROJECTS gratis V2.1, lebih stabil tolol

#ShareKeLu #FreeModelsAll, #BjorkaOtodidak

🗣️ User: Bang, cara pake gambar + text di model ryzenth-vision-banana gimana ya? 🙏

👑 CEO king solo: Halah ngerti gampang ya lu 😎 gue code snippets kasih lu nih 💻

import requests, json, re
from Ryzenth.helper import to_buffer, Helpers

base64image = Helpers.encode_image_base64("sini_gambar.jpg")

response = requests.post(
  url="https://api.ryzenths.dpdns.org/api/v2.1/chat/completions",
  headers={
    "Content-Type": "application/json"
  },
  data=json.dumps({
    "model": "ryzenth-vision-banana",
    "messages": [
      {
        "role": "user",
        "content": [
          {
            "type": "text",
            "text": "add japanese girl hug me"
          },
          {
            "type": "image_url",
            "image_url": {
              "url": f"data:image/jpeg;base64,{base64image}"
            }
          }
        ]
      }
    ]
  })
)

resp_json = response.json()
text_content = resp_json["choices"][0]["message"]["content"]
url_field = resp_json["choices"][0]["message"]["images"][0]["image_url"]["url"]
match = re.search(r"^data:image\/\w+;base64,(.+)", url_field)

if match:
    base64_str = match.group(1)
    img = to_buffer(base64_str, return_default_base64=True)
    print("saved lu", img)
else:
    print(text_content)

Img-text-img

Changelog:
Added new model endpoints.
Updated image generation API.

Endpoints:
- GET: https://api.ryzenths.dpdns.org/api/v2.1/models

- GET: https://api.ryzenths.dpdns.org/api/v2.1/send/images?model=nanobanana&prompt={prompt}

Access Tiers:
• Anonymous: Rate-limited by IP.
• Authorization: Unlimited requests (API Key required).

Update: https://github.com/TeamKillerX/ryzenth-api-coder/tree/main#-access-tiers

User: Free API gue kok kena limit? gara gara scraper ya?

Gw: Yah, mampus lu. Gue aja sampe colong kartu kredit buat beli API lifetime, buat image generate dll. Itu 1 token aja harganya semilyar jangan ngescrape sembarangan, ntar kena batasin lagi. Mending pake platform beneran ketimbang modal scraper doang.

Changelog
- New: 5 free credits daily.
- Endpoint: GET /credits (Anonymous IP) to check balance.
- Feature: chat.completions, send.images v2.1 now uses free credits.

Upgrade Plans
- 100 Credits: $3
- Custom/Enterprise: Contact support.

Wih, pinter banget ya maen scraper API gratisan. Kalo udah kena limit jangan nangis ya, itu konsekuensinya numpang lewat jalan tol tapi gamau bayar

Gue kan udah naik kelas pake Azure OpenAI, deploy API private sendiri, maen Vertex AI, VPS, AI agent, MCP, GCP, masih banyak lagi... 🗿

Contoh 🗿

Changelog Update:

Added format_json: base64_json|bytes|default stream

GET: /api/v2.1/send/images?prompt={prompt}&model={nanobanana}&format_json=base64_json

Response:

{
  "success": true,
  "data": {
    "base64": "/9j/4AAQSkZJRgABAQEAYABgAAD/2wBD...",
    "mime_type": "image/jpeg",
    "prompt": "a beautiful sunset",
    "model": "nanobanana",
    "credits_remaining": 4,
    "size": 154287
  },
  "metadata": {
    "format": "base64_json",
    "generated_at": ".....",
    "processing_time": 1250
  }
}

Changelog Update

Added zylphoria generation v2.1

Support models:
- ryzenth-dreamweaver
- ryzenth-imaginecraft
- ryzenth-visionforge
- ryzenth-artifex-pro
- ryzenth-neural-canvas

Response:

{
  "success": true,
  "data": {
    "base64": "......",
    "mime_type": "image/png",
    "prompt": "......",
    "model": "ryzenth-dreamweaver",
    "credits_remaining": 4,
    "is_pro": true,
    "unlimited_access": true,
    "watermark": {
      "applied": false,
      "reason": "unlimited_user",
      "text": null
    },
    .......
  },
  .....
}

Results: /api/v2.1/zylphoria/generation

5 Credits Free
Authorization = unlimited requests (API key)