Metadata-Version: 2.4
Name: sunwaee-gen
Version: 0.0.10
Summary: All LLMs, one response format.
Author: David NAISSE
Maintainer: David NAISSE
License: MIT License
        
        Copyright (c) 2025 DvdNss
        
        Permission is hereby granted, free of charge, to any person obtaining a copy
        of this software and associated documentation files (the "Software"), to deal
        in the Software without restriction, including without limitation the rights
        to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
        copies of the Software, and to permit persons to whom the Software is
        furnished to do so, subject to the following conditions:
        
        The above copyright notice and this permission notice shall be included in all
        copies or substantial portions of the Software.
        
        THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
        IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
        FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
        AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
        LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
        OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
        SOFTWARE.
        
Requires-Python: >=3.10
Description-Content-Type: text/markdown
License-File: LICENSE
Requires-Dist: aiohttp
Requires-Dist: fastapi[standard]
Requires-Dist: pydantic
Requires-Dist: rich
Requires-Dist: typer[all]
Requires-Dist: uvicorn[standard]
Provides-Extra: dev
Requires-Dist: pytest; extra == "dev"
Requires-Dist: pytest-cov; extra == "dev"
Requires-Dist: pytest-asyncio; extra == "dev"
Requires-Dist: httpx; extra == "dev"
Requires-Dist: twine; extra == "dev"
Requires-Dist: build; extra == "dev"
Dynamic: license-file

# SUNWÆE-GEN

All LLMs, one response format.

[![Python](https://img.shields.io/badge/python-3.10+-blue.svg)](https://python.org)
[![Tests](https://img.shields.io/badge/tests-118%20passed-brightgreen.svg)](tests/)
[![Coverage](https://img.shields.io/badge/coverage-100%25-brightgreen.svg)](tests/)
[![License](https://img.shields.io/badge/license-MIT-blue.svg)](LICENSE)

## OVERVIEW

All LLMs, one response format - available through SDK or self-hosted API. Includes usage, cost and performance metrics like reasoning duration and throughput (see [response format](#response)).

> GEN doesn't use provider-specific libraries (e.g. openai, anthropic, google...) and parses the raw HTTP responses (including server-sent event streams) directly from the providers using provider-specific logic called `adapters`.

What GEN does under the hood:

1. validates messages according to the openai format (see [MODELS](#models))
2. validates tools according to the openai format (see [MODELS](#models))
3. use provider-specific logic (`adapters`) to build provider-specific payload.
4. use provider-specific logic (`adapters`) to parse provider-specific response.
5. compute additional metrics related to performance, cost, usage... and return a [block](#response)

> Reasoning tokens are not available for certain agents despite supporting reasoning (e.g. `openai/gpt-5`, `google/gemini-2.5-pro`...). When that's the case, a block with `reasoning="reasoning started, but reasoning tokens are not avaiable for this model..."` will be returned, indicating reasoning has started.

## INSTALLATION

1. For regular usage:

```bash
pip install sunwaee-gen
```

2. For dev usage (contributors):

```bash
pip install "sunwaee-gen[dev]"
```

### USAGE

1. SDK

```python
import asyncio
import sunwaee_gen as gen

# list available resources
print(gen.AGENTS)
print(gen.MODELS)
print(gen.PROVIDERS)

messages = [
    {"role": "system": "content": "You are a helpful assistant."},
    {"role": "user", "content": "What's the latest news about AI?"}
]

tools = [
    {
        "type": "function",
        "function": {
            "name": "search_web",
            "description": "Search the web for information",
            "parameters": {
                "type": "object",
                "properties": {
                    "query": {"type": "string", "description": "Search query"}
                },
                "required": ["query"]
            }
        }
    }
]

async def main():

    # NOTE we use `async for` for both the regular and streaming completion
    async for block in gen.async_completion(
        "openai/gpt-5",
        messages=messages,
        tools=tools,
        streaming=False
    ):
        if block["reasoning"]:
            print(f"🤔 Reasoning: {block['reasoning']}")
        if block["content"]:
            print(f"💬 Content: {block['content']}")
        if block["tool_calls"]:
            print(f"🔧 Tool calls: {len(block['tool_calls'])}")

asyncio.run(main())
```

2. API

```sh
# serve
uvicorn sunwaee_gen.api:api

curl -X 'POST' \
  'http://localhost:8000/completion' \
  -H 'accept: application/json' \
  -H 'Authorization: Bearer <api_key>' \
  -H 'Content-Type: application/json' \
  -d '{
  "agent": "openai/gpt-5-nano",
  "messages": [{"role": "user", "content": "hi"}],
  "tools": [],
  "streaming": false
}'
```

### MODELS

#### MESSAGES

```python
[
    {
        "role": "system|user|assistant|tool",
        "content": "Here's a response.", # str
        "tool_call_id": "tc_123", # str
        "tool_calls": [
            {
            "id": "tc_123", # str
            "type": "function",
            "function": {
                "name": "get_weather", # str
                "arguments": "{\"city\": \"Paris\"}" # str
                }
            }
        ]
    }
]
```

> **`tool` messages must contain `tool_call_id`.**

#### TOOLS

```python
[
    {
        "type": "function", # str
        "function": {
            "name": "get_weather", # str
            "description": "Get a weather in a given city.", # str
            "parameters": {
                "type": "object", # str
                "properties": {
                    "city": {
                        "type": "string", # str
                        "description": "The city (e.g. Paris, London...)" # str
                    }
                },
                "required": ["city"] # list[str]
            }
        }
    }
]
```

#### RESPONSE

```python
{
  "model": {
    "name": "string",
    "display_name": "string",
    "origin": "string",
    "version": "string"
  },
  "provider": {
    "name": "string",
    "url": "string"
  },
  "error": {
    "status": 0
    "message": "string",
  },
  "usage": {
    "prompt_tokens": 0,
    "completion_tokens": 0,
    "total_tokens": 0
  },
  "cost": {
    "prompt_cost": 0,
    "completion_cost": 0,
    "total_cost": 0
  },
  "performance": {
    "latency": 0,
    "reasoning_duration": 0,
    "content_duration": 0,
    "total_duration": 0,
    "throughput": 0
  },
  "reasoning": "string",
  "content": "string",
  "tool_calls": [],
  "raw": "string",
  "streaming": false
}
```
