ramalama-chat(1)

📦
version 0.25.0 · 4 versions
, in fedora 43, fedora 44, nixos 26.05, nixpkgs unstable

✨ ramalama-chat ✨

okay bestie, gather round — this is the ramalama-chat page from the General Commands Manual, and it's literally about chatting with an OpenAI-compatible REST API from your terminal. main character energy fr fr 💅

🌸 NAME 🌸

ramalama-chat — OpenAI chat with the specified REST API URL 🦋

🎀 SYNOPSIS 🎀

ramalama chat [options] [arg...]

positional arguments:

  • ARGS — overrides the default prompt, and the output is returned without entering the chatbot (unless --interactive is specified) 👑

✨ DESCRIPTION ✨

So basically, you're chatting with an OpenAI REST API. By default, if you hand it arguments or pipe stuff in via stdin, it shows you the response and then dips out (exits). 🌈

But if you want to keep the conversation going, hit --interactive and it'll stay in an interactive chat session after it finishes processing that initial prompt. no cap 🫶

💖 OPTIONS 💖

  • --api-key — OpenAI-compatible API key 🔑. You can also set it with the RAMALAMA_API_KEY environment variable instead.
  • --color — whether or not to use color in the chat 🎨. Possible values are "never", "always" and "auto". (default: auto)
  • --help, -h — shows this help message and exits 🫡
  • --interactive, -i — keeps you in interactive chat mode after it processes stdin or prompt arguments 💬. By default, when arguments or piped input are provided, the command exits after displaying the response. This flag lets you keep chatting interactively instead of getting abandoned mid-convo 😭
  • --list — lists the available models at an endpoint 📋
  • --max-tokens=integer — maximum number of tokens to generate. Set it to 0 for unlimited output (default: 0). 🌊
  • --mcp=SERVER_URL — MCP (Model Context Protocol) servers to use for enhanced tool calling capabilities 🛠️. You can specify it multiple times to connect to multiple MCP servers. Each server provides tools that can be automatically invoked during chat conversations. lowkey iconic
  • --model=MODEL — model for inferencing 🧠 (may not be required for endpoints that only serve one model)
  • --prefix — prefix for the user prompt (default: 🦭 > ) 🦭
  • --rag=path — a file or directory of files to be loaded and provided as local context in the chat history 📚
  • --summarize-after=N — automatically summarize conversation history after N messages to prevent context growth 📉. When enabled, ramalama will periodically condense older messages into a summary, keeping only recent messages and the summary. This prevents the context from growing indefinitely during long chat sessions. Set to 0 to disable (default: 4).
  • --temp=float — temperature of the response from the AI Model 🌡️. Lower numbers are more deterministic, higher numbers are more creative. pick your vibe bestie
  • --url=URL — the host to send requests to (default: http://127.0.0.1:8080) 🌐

🧚 INTERACTIVE COMMANDS 🧚

When you're running in interactive chat mode, these commands are available. All commands are case-insensitive (e.g., /CLEAR, /Clear, and /clear all work). 💅

  • /help, help, ? — displays help information showing all available commands and their descriptions 🆘
  • /clear — clears the conversation history without exiting the chat session 🧹. This resets the context and lets you start a fresh conversation without restarting the container or connection. A confirmation message will be displayed when the history is cleared.
  • /bye, exit — exits the chat session and closes the connection 👋
  • /tool [question] — (Only available when using --mcp) manually select which MCP tool to use for a question 🛠️. Without this command, the AI automatically decides whether to use tools based on the question.
  • Ctrl + D — exits the chat session (EOF signal) 🚪

🔥 EXAMPLES 🔥

Communicate with the default local OpenAI REST API. (http://127.0.0.1:8080) With Podman containers. 🦭

$ ramalama chat
🦭 >

Communicate with an alternative OpenAI REST API URL. With Docker containers. 🐋

$ ramalama chat --url http://localhost:1234
🐋 >

Send multiple lines at once

$ ramalama chat
🦭 > Hi
🦭 > tell me a funny story
🦭 > please

Send an initial prompt via stdin and continue chatting

$ echo "What is 2+2?" | ramalama chat --interactive
4
🦭 > Now multiply that by 3
12
🦭 > /bye

Clear conversation history during a chat session (commands are case-insensitive)

$ ramalama chat
🦭 > What is 2+2?
4
🦭 > /CLEAR
Conversation history cleared.
🦭 >

🌈 SEE ALSO 🌈

💜 HISTORY 💜

Jun 2025, Originally compiled by Dan Walsh dwalsh@redhat.com ⟨mailto:dwalsh@redhat.com⟩ ✨

and that's the whole page bestie, ramalama-chat(1) from the General Commands Manual — go forth and chat, you absolute icon 👑🦋💖

📦 where this page comes from