What makes this unlimited AI chat
- →No per-message or per-token billing — one flat price for 24 hours or 30 days
- →No usage window that locks you out after a few hours of chatting
- →Switch between 17 models in the same chat app
- →Up to 1M-token context for long documents and codebases
- →Your plan sets how many replies run at the same time: 8 or 16
Open WebUI
- →Open Admin Panel → Settings → Connections
- →Under OpenAI API, add a connection
- →URL: https://api.unlimitedai.sh/v1
- →Key: your UnlimitedAI key (sk-xx)
- →Save — the models appear in the model picker
LibreChat
librechat.yaml
endpoints:
custom:
- name: "UnlimitedAI"
apiKey: "sk-xx"
baseURL: "https://api.unlimitedai.sh/v1"
models:
default: ["anthropic/claude-sonnet-5", "openai/gpt-6-luna", "deepseek/deepseek-v4-pro-0813"]
fetch: trueLobeChat
- →Open Settings → Language Model → OpenAI
- →API Key: sk-xx
- →API proxy URL: https://api.unlimitedai.sh/v1
- →Add the model IDs you want to the model list, e.g. anthropic/claude-opus-4.8
Chatbox and other chat apps
- →Add a custom provider of type “OpenAI API compatible”
- →API host / base URL: https://api.unlimitedai.sh/v1
- →API key: sk-xx
- →Model: any ID from the model list, e.g. z-ai/glm-5.3 or moonshotai/kimi-k3
- →Any chat app that accepts a custom OpenAI base URL works the same way
Which model to chat with
- →Best writing and reasoning: anthropic/claude-opus-4.8 or anthropic/claude-sonnet-5.5
- →Fast everyday answers: deepseek/deepseek-v4.1-flash or z-ai/glm-5.3-flash
- →Very long documents: moonshotai/kimi-k3 or qwen/qwen3.8-max
- →All 17 Models and Ultron only: anthropic/claude-opus-5.5, openai/gpt-6.1-sol, openai/gpt-6-astra
Good to know
- →UnlimitedAI does not host a chat website — you use your own chat app
- →UnlimitedAI is independent and not affiliated with Anthropic, OpenAI or the chat apps above
- →Keys are for one person; sharing a key with other users gets it revoked
Questions
- Is there an unlimited AI chat like ChatGPT?
- Yes — connect an UnlimitedAI key to a ChatGPT-style app such as Open WebUI, LibreChat, LobeChat or Chatbox. You chat with Claude, GPT, DeepSeek and more with no message caps or token counting.
- Do I need to code?
- No. In most chat apps you only paste a base URL, your key and a model name into the settings.
- What is the cheapest way to try it?
- The Max pass: $20 for 24 hours on 12 models. Ultron gives all 17 models for $35 for 24 hours.