# Groq Chat (https://docs.fuser.studio/docs/nodes/llm/groq-chat)

_Provided by [Groq](https://groq.com) — [Learn More](https://groq.com)._

## Inputs [#inputs]

- **Prompt** (`textarea`) — The message prompt to send to the model.

- **Model** (`dropdown`) — The Groq model to use for chat completion.

   Optional: `true` • Default: `openai/gpt-oss-120b` • Accepts: `openai/gpt-oss-20b`, `openai/gpt-oss-120b`, `qwen/qwen3-32b`, `llama-3.1-8b-instant`, `llama-3.3-70b-versatile`, `meta-llama/llama-guard-4-12b`, `meta-llama/llama-4-maverick-17b-128e-instruct`, `meta-llama/llama-4-scout-17b-16e-instruct`, `moonshotai/kimi-k2-instruct`

- **System Prompt** (`textarea`) — A system prompt that sets the behavior and context for the conversation.

   Optional: `true`

- **Attachment** (`string`) _(accepts multiple)_ — Attachments to include with the prompt (images, text, etc.).

   Optional: `true`

- **Last Message Only** (`toggle`) — If enabled, only the last message will be sent to the model instead of the full conversation history.

   Optional: `true` • Default: `false`

- **Temperature** (`number`) — Controls randomness in the output. Higher values make the output more random, lower values make it more deterministic.

   Optional: `true` • Minimum: `-1` • Maximum: `1`

- **Seed** (`seed`) — A seed value for reproducible outputs. Use the same seed to get consistent results.

   Optional: `true` • Default: `8089`

- **Top P** (`number`) — Nucleus sampling parameter. Controls diversity via nucleus sampling.

   Optional: `true`

- **Top K** (`number`) — Limits the number of highest probability tokens to consider.

   Optional: `true`

- **Max Tokens** (`number`) — Maximum number of tokens to generate in the response.

   Optional: `true`

- **Stop Sequences** (`string`) _(accepts multiple)_ — Sequences that will cause the model to stop generating further tokens.

   Optional: `true`

- **Presence Penalty** (`number`) — Penalizes tokens based on whether they appear in the text so far.

   Optional: `true` • Minimum: `-2` • Maximum: `2`

- **Frequency Penalty** (`number`) — Penalizes tokens based on their frequency in the text so far.

   Optional: `true` • Minimum: `-2` • Maximum: `2`

## Outputs [#outputs]

- **Last Response** (`textarea`) — The last response from the model.