neom_ollama 1.2.0

Ollama integration for Dart — model discovery, pull/delete, health checks, setup automation, OpenAI-compatible chat, hardware profiling, reasoning trace parsing, and plain-text tool-call recovery for [...]

neom_ollama

Ollama integration for Dart and Flutter. Model discovery, pull/delete, health checks, setup automation, and OpenAI-compatible chat service for local inference.

Part of the Open Neom ecosystem.

Features

• OllamaClient — Full Ollama REST API: status, list/show/pull/create/delete models, chat, streaming.
• OllamaSetup — Automated provisioning: detect install, start server, pull models with progress.
• OllamaChatService — OpenAI-compatible /v1/chat/completions client with conversation history.
• Pure Dart — No Flutter dependency. Works in CLI, desktop, mobile, and server.

Usage

import 'package:neom_ollama/neom_ollama.dart';

// Check if Ollama is running
final client = OllamaClient();
final status = await client.checkStatus(); // OllamaStatus.running

// List local models
final models = await client.listModels();
for (final m in models) {
  print('${m.displayName} — ${m.sizeLabel}');
}

// Single-turn chat
final response = await client.chat('qwen2.5:3b', 'What is Dart?');
print(response);

// Streaming chat
await for (final chunk in client.chatStream('qwen2.5:3b', 'Explain async/await')) {
  stdout.write(chunk);
}

// Pull a model with progress
await for (final p in client.pullModel('llama3.1:8b')) {
  print('${p.status} ${((p.progress ?? 0) * 100).toStringAsFixed(0)}%');
}

// Create a custom model from Modelfile
await client.createModel('my-model', 'FROM qwen2.5:3b\nSYSTEM "You are helpful."');

// OpenAI-compatible chat service
final chat = OllamaChatService(model: 'qwen2.5:3b');
chat.setSystemInstruction('You are a coding assistant.');
final reply = await chat.sendMessage('Write a hello world in Dart');
print(reply);

Automated Setup

final setup = OllamaSetup();

// Check status
final status = await setup.checkStatus(requiredModel: 'qwen2.5:3b');

// Full automated setup with progress
await setup.runFullSetup(
  modelName: 'qwen2.5:3b',
  onProgress: (step, message, progress) {
    print('[$step] $message (${(progress * 100).toStringAsFixed(0)}%)');
  },
);

Requirements

• Ollama installed and running on localhost:11434
• Dart SDK ≥ 3.8.0

License

Apache 2.0 — see LICENSE.