local-openai2anthropic 0.1.0__py3-none-any.whl → 0.3.6__py3-none-any.whl

local_openai2anthropic-0.3.6.dist-info/METADATA

@@ -0,0 +1,374 @@
+Metadata-Version: 2.4
+Name: local-openai2anthropic
+Version: 0.3.6
+Summary: A lightweight proxy server that converts Anthropic Messages API to OpenAI API
+Project-URL: Homepage, https://github.com/dongfangzan/local-openai2anthropic
+Project-URL: Repository, https://github.com/dongfangzan/local-openai2anthropic
+Project-URL: Issues, https://github.com/dongfangzan/local-openai2anthropic/issues
+Author-email: dongfangzan <zsybook0124@163.com>
+Maintainer-email: dongfangzan <zsybook0124@163.com>
+License: Apache-2.0
+License-File: LICENSE
+Keywords: anthropic,api,claude,messages,openai,proxy
+Classifier: Development Status :: 4 - Beta
+Classifier: Intended Audience :: Developers
+Classifier: License :: OSI Approved :: Apache Software License
+Classifier: Programming Language :: Python :: 3
+Classifier: Programming Language :: Python :: 3.12
+Classifier: Programming Language :: Python :: 3.13
+Classifier: Topic :: Software Development :: Libraries :: Python Modules
+Requires-Python: >=3.12
+Requires-Dist: anthropic>=0.30.0
+Requires-Dist: fastapi>=0.100.0
+Requires-Dist: httpx>=0.25.0
+Requires-Dist: openai>=1.30.0
+Requires-Dist: pydantic-settings>=2.0.0
+Requires-Dist: pydantic>=2.0.0
+Requires-Dist: tomli>=2.0.0; python_version < '3.11'
+Requires-Dist: uvicorn[standard]>=0.23.0
+Provides-Extra: dev
+Requires-Dist: black>=23.0.0; extra == 'dev'
+Requires-Dist: mypy>=1.0.0; extra == 'dev'
+Requires-Dist: pytest-asyncio>=0.21.0; extra == 'dev'
+Requires-Dist: pytest>=7.0.0; extra == 'dev'
+Requires-Dist: ruff>=0.1.0; extra == 'dev'
+Description-Content-Type: text/markdown
+
+# local-openai2anthropic
+
+[![Python 3.12+](https://img.shields.io/badge/python-3.12+-blue.svg)](https://www.python.org/downloads/)
+[![License: Apache 2.0](https://img.shields.io/badge/License-Apache%202.0-blue.svg)](https://opensource.org/licenses/Apache-2.0)
+[![PyPI](https://img.shields.io/pypi/v/local-openai2anthropic.svg)](https://pypi.org/project/local-openai2anthropic/)
+
+**English | [中文](README_zh.md)**
+
+A lightweight proxy that lets applications built with [Claude SDK](https://github.com/anthropics/anthropic-sdk-python) talk to locally-hosted OpenAI-compatible LLMs.
+
+---
+
+## What Problem This Solves
+
+Many local LLM tools (vLLM, SGLang, etc.) provide an OpenAI-compatible API. But if you've built your app using Anthropic's Claude SDK, you can't use them directly.
+
+This proxy translates Claude SDK calls to OpenAI API format in real-time, enabling:
+
+- **Local LLM inference** with Claude-based apps
+- **Offline development** without cloud API costs
+- **Privacy-first AI** - data never leaves your machine
+- **Seamless model switching** between cloud and local
+- **Web Search tool** - built-in Tavily web search for local models
+
+---
+
+## Supported Local Backends
+
+Currently tested and supported:
+
+| Backend | Description | Status |
+|---------|-------------|--------|
+| [vLLM](https://github.com/vllm-project/vllm) | High-throughput LLM inference | ✅ Fully supported |
+| [SGLang](https://github.com/sgl-project/sglang) | Fast structured language model serving | ✅ Fully supported |
+
+Other OpenAI-compatible backends may work but are not fully tested.
+
+---
+
+## Quick Start
+
+### 1. Install
+
+```bash
+pip install local-openai2anthropic
+```
+
+### 2. Configure Your LLM Backend (Optional)
+
+**Option A: Start a local LLM server**
+
+If you don't have an LLM server running, you can start one locally:
+
+Example with vLLM:
+```bash
+vllm serve meta-llama/Llama-2-7b-chat-hf
+# vLLM starts OpenAI-compatible API at http://localhost:8000/v1
+```
+
+Or with SGLang:
+```bash
+sglang launch --model-path meta-llama/Llama-2-7b-chat-hf --port 8000
+# SGLang starts at http://localhost:8000/v1
+```
+
+**Option B: Use an existing OpenAI-compatible API**
+
+If you already have a deployed OpenAI-compatible API (local or remote), you can use it directly. Just note the base URL for the next step.
+
+Examples:
+- Local vLLM/SGLang: `http://localhost:8000/v1`
+- Remote API: `https://api.example.com/v1`
+
+> **Note:** If you're using [Ollama](https://ollama.com), it natively supports the Anthropic API format, so you don't need this proxy. Just point your Claude SDK directly to `http://localhost:11434/v1`.
+
+### 3. Start the Proxy
+
+**Option A: Run in background (recommended)**
+
+```bash
+export OA2A_OPENAI_BASE_URL=http://localhost:8000/v1  # Your local LLM endpoint
+export OA2A_OPENAI_API_KEY=dummy  # Any value, not used by local backends
+
+oa2a start              # Start server in background
+# Server starts at http://localhost:8080
+
+# View logs
+oa2a logs               # Show last 50 lines of logs
+oa2a logs -f            # Follow logs in real-time (Ctrl+C to exit)
+
+# Check status
+oa2a status             # Check if server is running
+
+# Stop server
+oa2a stop               # Stop background server
+
+# Restart server
+oa2a restart            # Restart with same settings
+```
+
+**Option B: Run in foreground**
+
+```bash
+export OA2A_OPENAI_BASE_URL=http://localhost:8000/v1
+export OA2A_OPENAI_API_KEY=dummy
+
+oa2a                    # Run server in foreground (blocking)
+# Press Ctrl+C to stop
+```
+
+### 4. Use in Your App
+
+```python
+import anthropic
+
+client = anthropic.Anthropic(
+    base_url="http://localhost:8080",  # Point to proxy
+    api_key="dummy-key",  # Not used
+)
+
+message = client.messages.create(
+    model="meta-llama/Llama-2-7b-chat-hf",  # Your local model name
+    max_tokens=1024,
+    messages=[{"role": "user", "content": "Hello!"}],
+)
+
+print(message.content[0].text)
+```
+
+---
+
+## Using with Claude Code
+
+You can configure [Claude Code](https://github.com/anthropics/claude-code) to use your local LLM through this proxy.
+
+### Configuration Steps
+
+1. **Edit Claude Code config file** at `~/.claude/settings.json`:
+
+```json
+{
+  "env": {
+    "ANTHROPIC_BASE_URL": "http://localhost:8080",
+    "ANTHROPIC_API_KEY": "dummy-key",
+    "ANTHROPIC_MODEL": "meta-llama/Llama-2-7b-chat-hf",
+    "ANTHROPIC_DEFAULT_SONNET_MODEL": "meta-llama/Llama-2-7b-chat-hf",
+    "ANTHROPIC_DEFAULT_OPUS_MODEL": "meta-llama/Llama-2-7b-chat-hf",
+    "ANTHROPIC_DEFAULT_HAIKU_MODEL": "meta-llama/Llama-2-7b-chat-hf",
+    "ANTHROPIC_REASONING_MODEL": "meta-llama/Llama-2-7b-chat-hf"
+  }
+}
+```
+
+| Variable | Description |
+|----------|-------------|
+| `ANTHROPIC_MODEL` | General model setting |
+| `ANTHROPIC_DEFAULT_SONNET_MODEL` | Default model for Sonnet mode (Claude Code default) |
+| `ANTHROPIC_DEFAULT_OPUS_MODEL` | Default model for Opus mode |
+| `ANTHROPIC_DEFAULT_HAIKU_MODEL` | Default model for Haiku mode |
+| `ANTHROPIC_REASONING_MODEL` | Default model for reasoning tasks |
+
+2. **Or set environment variables** before running Claude Code:
+
+```bash
+export ANTHROPIC_BASE_URL=http://localhost:8080
+export ANTHROPIC_API_KEY=dummy-key
+
+claude
+```
+
+### Complete Workflow Example
+
+Make sure `~/.claude/settings.json` is configured as described above.
+
+Terminal 1 - Start your local LLM:
+```bash
+vllm serve meta-llama/Llama-2-7b-chat-hf
+```
+
+Terminal 2 - Start the proxy (background mode):
+```bash
+export OA2A_OPENAI_BASE_URL=http://localhost:8000/v1
+export OA2A_OPENAI_API_KEY=dummy
+export OA2A_TAVILY_API_KEY="tvly-your-tavily-api-key"  # Optional: enable web search
+
+oa2a start
+```
+
+Terminal 3 - Launch Claude Code:
+```bash
+claude
+```
+
+Now Claude Code will use your local LLM instead of the cloud API.
+
+To stop the proxy:
+```bash
+oa2a stop
+```
+
+---
+
+## Features
+
+- ✅ **Streaming responses** - Real-time token streaming via SSE
+- ✅ **Tool calling** - Local LLM function calling support
+- ✅ **Vision models** - Multi-modal input for vision-capable models
+- ✅ **Web Search** - Give your local LLM internet access (see below)
+- ✅ **Thinking mode** - Supports reasoning/thinking model outputs
+
+---
+
+## Web Search Capability 🔍
+
+**Bridge the gap: Give your local LLM the web search power that Claude Code users enjoy!**
+
+When using locally-hosted models with Claude Code, you lose access to the built-in web search tool. This proxy fills that gap by providing a server-side web search implementation powered by [Tavily](https://tavily.com).
+
+### The Problem
+
+| Scenario | Web Search Available? |
+|----------|----------------------|
+| Using Claude (cloud) in Claude Code | ✅ Built-in |
+| Using local vLLM/SGLang in Claude Code | ❌ Not available |
+| **Using this proxy + local LLM** | ✅ **Enabled via Tavily** |
+
+### How It Works
+
+```
+Claude Code → Anthropic SDK → This Proxy → Local LLM
+                                      ↓
+                                 Tavily API (Web Search)
+```
+
+The proxy intercepts `web_search_20250305` tool calls and handles them directly, regardless of whether your local model supports web search natively.
+
+### Setup Tavily Search
+
+1. **Get a free API key** at [tavily.com](https://tavily.com) - generous free tier available
+
+2. **Configure the proxy:**
+```bash
+export OA2A_OPENAI_BASE_URL=http://localhost:8000/v1
+export OA2A_OPENAI_API_KEY=dummy
+export OA2A_TAVILY_API_KEY="tvly-your-tavily-api-key"  # Enable web search
+
+oa2a
+```
+
+3. **Use in your app:**
+```python
+import anthropic
+
+client = anthropic.Anthropic(
+    base_url="http://localhost:8080",
+    api_key="dummy-key",
+)
+
+message = client.messages.create(
+    model="meta-llama/Llama-2-7b-chat-hf",
+    max_tokens=1024,
+    tools=[
+        {
+            "name": "web_search_20250305",
+            "description": "Search the web for current information",
+            "input_schema": {
+                "type": "object",
+                "properties": {
+                    "query": {"type": "string", "description": "Search query"},
+                },
+                "required": ["query"],
+            },
+        }
+    ],
+    messages=[{"role": "user", "content": "What happened in AI today?"}],
+)
+
+if message.stop_reason == "tool_use":
+    tool_use = message.content[-1]
+    print(f"Searching: {tool_use.input}")
+    # The proxy automatically calls Tavily and returns results
+```
+
+### Tavily Configuration Options
+
+| Variable | Default | Description |
+|----------|---------|-------------|
+| `OA2A_TAVILY_API_KEY` | - | Your Tavily API key ([get free at tavily.com](https://tavily.com)) |
+| `OA2A_TAVILY_MAX_RESULTS` | 5 | Number of search results to return |
+| `OA2A_TAVILY_TIMEOUT` | 30 | Search timeout in seconds |
+| `OA2A_WEBSEARCH_MAX_USES` | 5 | Max search calls per request |
+
+---
+
+## Configuration
+
+| Variable | Required | Default | Description |
+|----------|----------|---------|-------------|
+| `OA2A_OPENAI_BASE_URL` | ✅ | - | Your local LLM's OpenAI-compatible endpoint |
+| `OA2A_OPENAI_API_KEY` | ✅ | - | Any value (local backends usually ignore this) |
+| `OA2A_PORT` | ❌ | 8080 | Proxy server port |
+| `OA2A_HOST` | ❌ | 0.0.0.0 | Proxy server host |
+| `OA2A_TAVILY_API_KEY` | ❌ | - | Enable web search ([tavily.com](https://tavily.com)) |
+
+---
+
+## Architecture
+
+```
+Your App (Claude SDK)
+         │
+         ▼
+┌─────────────────────┐
+│  local-openai2anthropic  │  ← This proxy
+│  (Port 8080)        │
+└─────────────────────┘
+         │
+         ▼
+Your Local LLM Server
+(vLLM / SGLang)
+(OpenAI-compatible API)
+```
+
+---
+
+## Development
+
+```bash
+git clone https://github.com/dongfangzan/local-openai2anthropic.git
+cd local-openai2anthropic
+pip install -e ".[dev]"
+
+pytest
+```
+
+## License
+
+Apache License 2.0

local_openai2anthropic-0.3.6.dist-info/RECORD

@@ -0,0 +1,25 @@
+local_openai2anthropic/__init__.py,sha256=YEz1wpAzYlPY-zbmlQuLf8gpwLEVJBqff4LdfWcz6NM,1059
+local_openai2anthropic/__main__.py,sha256=K21u5u7FN8-DbO67TT_XDF0neGqJeFrVNkteRauCRQk,179
+local_openai2anthropic/config.py,sha256=y40uEMBE57dOGCV3w3v5j82ZPZZJWUnJ4yaFZXJ8pRk,4706
+local_openai2anthropic/converter.py,sha256=og94I514M9km_Wbk9c1ddU6fyaQNEbpd2zfpfnBQaTQ,16029
+local_openai2anthropic/daemon.py,sha256=pZnRojGFcuIpR8yLDNjV-b0LJRBVhgRAa-dKeRRse44,10017
+local_openai2anthropic/daemon_runner.py,sha256=rguOH0PgpbjqNsKYei0uCQX8JQOQ1wmtQH1CtW95Dbw,3274
+local_openai2anthropic/main.py,sha256=3xrjsKFBYK6B8niAtQz0U_yz-eTpf91HnHeAiR9CLQE,12174
+local_openai2anthropic/openai_types.py,sha256=jFdCvLwtXYoo5gGRqOhbHQcVaxcsxNnCP_yFPIv7rG4,3823
+local_openai2anthropic/protocol.py,sha256=VW3B1YrbYg5UAo7PveQv0Ny5vfuNa6yG6IlHtkuyXiI,5178
+local_openai2anthropic/router.py,sha256=gwSGCYQGd0tAj4B4cl30UDkIJDIfBP4D8T9KEMKnxyk,16196
+local_openai2anthropic/tavily_client.py,sha256=QsBhnyF8BFWPAxB4XtWCCpHCquNL5SW93-zjTTi4Meg,3774
+local_openai2anthropic/server_tools/__init__.py,sha256=QlJfjEta-HOCtLe7NaY_fpbEKv-ZpInjAnfmSqE9tbk,615
+local_openai2anthropic/server_tools/base.py,sha256=pNFsv-jSgxVrkY004AHAcYMNZgVSO8ZOeCzQBUtQ3vU,5633
+local_openai2anthropic/server_tools/web_search.py,sha256=1C7lX_cm-tMaN3MsCjinEZYPJc_Hj4yAxYay9h8Zbvs,6543
+local_openai2anthropic/streaming/__init__.py,sha256=RFKYQnc0zlhWK-Dm7GZpmabmszbZhY5NcXaaSsQ7Sys,227
+local_openai2anthropic/streaming/handler.py,sha256=X8viml6b40p-vr-A4HlEi5iCqmTsIMyQgj3S2RfweVE,22033
+local_openai2anthropic/tools/__init__.py,sha256=OM_6YAwy3G1kbrF7n5NvmBwWPGO0hwq4xLrYZFMHANA,318
+local_openai2anthropic/tools/handler.py,sha256=SO8AmEUfNIg16s6jOKBaYdajYc0fiI8ciOoiKXIJe_c,14106
+local_openai2anthropic/utils/__init__.py,sha256=0Apd3lQCmWpQHol4AfjtQe6A3Cpex9Zn-8dyK_FU8Z0,372
+local_openai2anthropic/utils/tokens.py,sha256=TV3vGAjoGZeyo1xPvwb5jto43p1U1f4HteCApB86X0g,3187
+local_openai2anthropic-0.3.6.dist-info/METADATA,sha256=B5M75TvhwturteqT_zxAJ7lXSTAy2QpjTCeDvPRixhQ,11293
+local_openai2anthropic-0.3.6.dist-info/WHEEL,sha256=WLgqFyCfm_KASv4WHyYy0P3pM_m7J5L9k2skdKLirC8,87
+local_openai2anthropic-0.3.6.dist-info/entry_points.txt,sha256=hdc9tSJUNxyNLXcTYye5SuD2K0bEQhxBhGnWTFup6ZM,116
+local_openai2anthropic-0.3.6.dist-info/licenses/LICENSE,sha256=X3_kZy3lJvd_xp8IeyUcIAO2Y367MXZc6aaRx8BYR_s,11369
+local_openai2anthropic-0.3.6.dist-info/RECORD,,