ollama-as-webapi

This repository provides a simple CLI tool to serve Ollama models as a Web API.

Prerequisites

• Node.js (v20 or higher recommended)
• Ollama installed and running

Installation

npm i -g @coder-ka/ollama-as-webapi

Usage

ollama-as-webapi

CLI Options

| Option | Description | Default | | --- | --- | --- | | --model | Model name | gemma3:4b | | --port | Port number | 3000 | | --host | Host name | localhost | | --gpu | Use GPU | true |

Web API Reference

Content Generation

/generate

Request Body

| Field | Required | Type | Description | Default | | --- | --- | --- | --- | --- | | prompt | Yes | string | Prompt to generate content for | "" | | maxTokens | No | number | Maximum number of tokens to generate | 256 | | temperature | No | number | Temperature for sampling | 0.7 | | topP | No | number | Top-p sampling | 0.9 | | topK | No | number | Top-k sampling | 40 | | n | No | number | Number of samples to generate | 1 | | stop | No | string[] | Stop sequences | [] |

Response Body

| Field | Type | Description | | --- | --- | --- | | text | string | Generated content | | tokens | number | Number of tokens generated | | finishReason | string | Reason for finishing generation |

Example

curl -X POST http://localhost:3000/generate \
  -H "Content-Type: application/json" \
  -d '{"prompt": "Hello"}'

License

MIT

How to contribute

1. Fork the repository
2. Create a feature branch
3. Commit your changes
4. Push to the branch
5. Open a Pull Request

Author

Katsuyuki Oeda