alltoken 0.2.0__tar.gz

alltoken-0.2.0/.gitignore

@@ -0,0 +1,46 @@
+# Python
+__pycache__/
+*.py[cod]
+*$py.class
+*.so
+*.egg
+*.egg-info/
+.eggs/
+
+# Build / dist
+build/
+dist/
+wheels/
+*.whl
+
+# Virtualenvs
+.venv/
+venv/
+env/
+ENV/
+
+# Test / coverage
+.pytest_cache/
+.coverage
+.coverage.*
+htmlcov/
+.tox/
+.nox/
+coverage.xml
+
+# Type checkers
+.mypy_cache/
+.pyright/
+.pytype/
+
+# Linters
+.ruff_cache/
+
+# Env
+.env
+.env.local
+
+# OS / editor
+.DS_Store
+.vscode/
+.idea/

alltoken-0.2.0/LICENSE

@@ -0,0 +1,21 @@
+MIT License
+
+Copyright (c) 2026 AllToken
+
+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.

alltoken-0.2.0/PKG-INFO

@@ -0,0 +1,118 @@
+Metadata-Version: 2.4
+Name: alltoken
+Version: 0.2.0
+Summary: Official Python SDK for AllToken — one API for OpenAI, Anthropic, and 100+ models.
+Project-URL: Homepage, https://alltoken.ai
+Project-URL: Repository, https://github.com/alltoken-ai/alltoken-python
+Project-URL: Documentation, https://alltoken.ai/docs
+Project-URL: Issues, https://github.com/alltoken-ai/alltoken-python/issues
+Author-email: AllToken <dev@alltoken.ai>
+License: MIT
+License-File: LICENSE
+Keywords: ai,alltoken,anthropic,api,llm,openai
+Classifier: Development Status :: 3 - Alpha
+Classifier: Intended Audience :: Developers
+Classifier: License :: OSI Approved :: MIT License
+Classifier: Programming Language :: Python :: 3
+Classifier: Programming Language :: Python :: 3.10
+Classifier: Programming Language :: Python :: 3.11
+Classifier: Programming Language :: Python :: 3.12
+Classifier: Programming Language :: Python :: 3.13
+Classifier: Topic :: Software Development :: Libraries :: Python Modules
+Requires-Python: >=3.10
+Requires-Dist: httpx>=0.27.0
+Requires-Dist: pydantic>=2.0.0
+Provides-Extra: dev
+Requires-Dist: datamodel-code-generator[ruff]>=0.26.0; extra == 'dev'
+Requires-Dist: mypy>=1.11.0; extra == 'dev'
+Requires-Dist: pytest>=8.0.0; extra == 'dev'
+Requires-Dist: ruff>=0.6.0; extra == 'dev'
+Description-Content-Type: text/markdown
+
+# alltoken-ai
+
+Official Python SDK for [AllToken](https://alltoken.ai) — one API for OpenAI, Anthropic, and 100+ models.
+
+```bash
+pip install alltoken-ai
+```
+
+Requires **Python 3.10+**.
+
+## Quick start
+
+```python
+from alltoken import AllToken
+
+client = AllToken(api_key="...")  # or os.environ["ALLTOKEN_API_KEY"]
+
+# OpenAI-compatible surface (maps to /v1)
+resp = client.openai.raw.post(
+    "/chat/completions",
+    json={
+        "model": "gpt-4o",
+        "messages": [{"role": "user", "content": "Hello!"}],
+    },
+)
+print(resp.json())
+
+# Anthropic-compatible surface (maps to /anthropic)
+resp = client.anthropic.raw.post(
+    "/messages",
+    json={
+        "model": "claude-sonnet-4",
+        "max_tokens": 1024,
+        "messages": [{"role": "user", "content": "Hello!"}],
+    },
+)
+print(resp.json())
+```
+
+The same API key works for both surfaces. Model catalog: [alltoken.ai/models](https://alltoken.ai/models).
+
+## Configuration
+
+```python
+AllToken(
+    api_key="...",                      # required
+    base_url="https://api.alltoken.ai", # optional, defaults to production
+    default_headers={"X-My-Tag": "a"},  # optional, merged into every request
+)
+```
+
+## API surface
+
+| Field | Spec | Base URL |
+|---|---|---|
+| `client.openai.raw` | `chat.yml` (OpenAI-compatible) | `https://api.alltoken.ai/v1` |
+| `client.anthropic.raw` | `anthropic.yml` | `https://api.alltoken.ai/anthropic` |
+
+`.raw` is a pre-configured [httpx.Client](https://www.python-httpx.org/api/#client) — base URL + auth are set, call `.get()` / `.post()` / `.stream()` directly. Pydantic models for request/response bodies are generated from the OpenAPI specs into `alltoken.generated.chat` and `alltoken.generated.anthropic`.
+
+## Status
+
+**v0.1.0 — Scaffold.** Pydantic models are generated from the spec, the wrapper surface is minimal. Expect breaking changes in 0.x. Ergonomic helpers (`client.chat.completions.create(...)`, async streaming iterators, retries, etc.) are coming in 0.2.x.
+
+## Contributing / Local development
+
+```bash
+# Clone megaopenrouter as a sibling (for the OpenAPI specs)
+git clone git@gitlab.53site.com:ai-innovation-lab/megaopenrouter.git ../megaopenrouter
+
+# Install with dev deps
+pip install -e ".[dev]"
+
+# Regenerate pydantic models from specs
+python scripts/generate.py
+
+# Test + lint
+pytest
+ruff check
+mypy .
+```
+
+Generated models live in `src/alltoken/generated/{chat,anthropic}.py` — these are **committed** so users who install from PyPI don't need to run codegen.
+
+## License
+
+[MIT](./LICENSE)

alltoken-0.2.0/README.md

@@ -0,0 +1,87 @@
+# alltoken-ai
+
+Official Python SDK for [AllToken](https://alltoken.ai) — one API for OpenAI, Anthropic, and 100+ models.
+
+```bash
+pip install alltoken-ai
+```
+
+Requires **Python 3.10+**.
+
+## Quick start
+
+```python
+from alltoken import AllToken
+
+client = AllToken(api_key="...")  # or os.environ["ALLTOKEN_API_KEY"]
+
+# OpenAI-compatible surface (maps to /v1)
+resp = client.openai.raw.post(
+    "/chat/completions",
+    json={
+        "model": "gpt-4o",
+        "messages": [{"role": "user", "content": "Hello!"}],
+    },
+)
+print(resp.json())
+
+# Anthropic-compatible surface (maps to /anthropic)
+resp = client.anthropic.raw.post(
+    "/messages",
+    json={
+        "model": "claude-sonnet-4",
+        "max_tokens": 1024,
+        "messages": [{"role": "user", "content": "Hello!"}],
+    },
+)
+print(resp.json())
+```
+
+The same API key works for both surfaces. Model catalog: [alltoken.ai/models](https://alltoken.ai/models).
+
+## Configuration
+
+```python
+AllToken(
+    api_key="...",                      # required
+    base_url="https://api.alltoken.ai", # optional, defaults to production
+    default_headers={"X-My-Tag": "a"},  # optional, merged into every request
+)
+```
+
+## API surface
+
+| Field | Spec | Base URL |
+|---|---|---|
+| `client.openai.raw` | `chat.yml` (OpenAI-compatible) | `https://api.alltoken.ai/v1` |
+| `client.anthropic.raw` | `anthropic.yml` | `https://api.alltoken.ai/anthropic` |
+
+`.raw` is a pre-configured [httpx.Client](https://www.python-httpx.org/api/#client) — base URL + auth are set, call `.get()` / `.post()` / `.stream()` directly. Pydantic models for request/response bodies are generated from the OpenAPI specs into `alltoken.generated.chat` and `alltoken.generated.anthropic`.
+
+## Status
+
+**v0.1.0 — Scaffold.** Pydantic models are generated from the spec, the wrapper surface is minimal. Expect breaking changes in 0.x. Ergonomic helpers (`client.chat.completions.create(...)`, async streaming iterators, retries, etc.) are coming in 0.2.x.
+
+## Contributing / Local development
+
+```bash
+# Clone megaopenrouter as a sibling (for the OpenAPI specs)
+git clone git@gitlab.53site.com:ai-innovation-lab/megaopenrouter.git ../megaopenrouter
+
+# Install with dev deps
+pip install -e ".[dev]"
+
+# Regenerate pydantic models from specs
+python scripts/generate.py
+
+# Test + lint
+pytest
+ruff check
+mypy .
+```
+
+Generated models live in `src/alltoken/generated/{chat,anthropic}.py` — these are **committed** so users who install from PyPI don't need to run codegen.
+
+## License
+
+[MIT](./LICENSE)

alltoken-0.2.0/alias/README.md

@@ -0,0 +1,11 @@
+# alltoken-ai
+
+This is an alias package that installs [alltoken](https://pypi.org/project/alltoken/), the official Python SDK for [AllToken](https://alltoken.ai).
+
+```bash
+pip install alltoken-ai
+# or equivalently:
+pip install alltoken
+```
+
+See the [alltoken package](https://pypi.org/project/alltoken/) for full documentation.

alltoken-0.2.0/pyproject.toml

@@ -0,0 +1,60 @@
+[build-system]
+requires = ["hatchling"]
+build-backend = "hatchling.build"
+
+[project]
+name = "alltoken"
+version = "0.2.0"
+description = "Official Python SDK for AllToken — one API for OpenAI, Anthropic, and 100+ models."
+readme = "README.md"
+license = { text = "MIT" }
+requires-python = ">=3.10"
+authors = [{ name = "AllToken", email = "dev@alltoken.ai" }]
+keywords = ["alltoken", "openai", "anthropic", "llm", "ai", "api"]
+classifiers = [
+  "Development Status :: 3 - Alpha",
+  "Intended Audience :: Developers",
+  "License :: OSI Approved :: MIT License",
+  "Programming Language :: Python :: 3",
+  "Programming Language :: Python :: 3.10",
+  "Programming Language :: Python :: 3.11",
+  "Programming Language :: Python :: 3.12",
+  "Programming Language :: Python :: 3.13",
+  "Topic :: Software Development :: Libraries :: Python Modules",
+]
+dependencies = [
+  "httpx>=0.27.0",
+  "pydantic>=2.0.0",
+]
+
+[project.urls]
+Homepage = "https://alltoken.ai"
+Repository = "https://github.com/alltoken-ai/alltoken-python"
+Documentation = "https://alltoken.ai/docs"
+Issues = "https://github.com/alltoken-ai/alltoken-python/issues"
+
+[project.optional-dependencies]
+dev = [
+  "datamodel-code-generator[ruff]>=0.26.0",
+  "pytest>=8.0.0",
+  "ruff>=0.6.0",
+  "mypy>=1.11.0",
+]
+
+[tool.hatch.build.targets.wheel]
+packages = ["src/alltoken"]
+
+[tool.hatch.build.targets.sdist]
+include = ["src/alltoken", "README.md", "LICENSE"]
+
+[tool.ruff]
+line-length = 100
+target-version = "py310"
+
+[tool.ruff.lint]
+select = ["E", "F", "I", "N", "UP", "B", "A", "C4", "SIM"]
+
+[tool.mypy]
+python_version = "3.10"
+strict = true
+files = ["src/alltoken"]

alltoken-0.2.0/src/alltoken/__init__.py

@@ -0,0 +1,63 @@
+"""alltoken-ai — Official Python SDK for AllToken.
+
+One API for OpenAI, Anthropic, and 100+ models.
+
+Example:
+    >>> from alltoken import AllToken
+    >>> client = AllToken(api_key="...")
+    >>> resp = client.openai.raw.post(
+    ...     "/chat/completions",
+    ...     json={"model": "gpt-4o", "messages": [{"role": "user", "content": "Hi"}]},
+    ... )
+    >>> print(resp.json())
+"""
+from __future__ import annotations
+
+from ._anthropic import AnthropicClient
+from ._client import AllTokenConfig
+from ._error import AllTokenError
+from ._openai import OpenAIClient
+from ._streaming import AllTokenStream
+from ._types import ChatCompletion, ChatCompletionChunk, ChatCompletionMessage
+
+__version__ = "0.1.0"
+
+
+class AllToken:
+    """Unified client exposing both OpenAI-compatible and Anthropic-compatible surfaces.
+
+    Both sub-clients share the same API key and base URL; only the path suffix differs
+    (/v1 vs /anthropic).
+    """
+
+    openai: OpenAIClient
+    anthropic: AnthropicClient
+
+    def __init__(
+        self,
+        *,
+        api_key: str,
+        base_url: str = "https://api.alltoken.ai",
+        default_headers: dict[str, str] | None = None,
+    ) -> None:
+        config = AllTokenConfig(
+            api_key=api_key,
+            base_url=base_url,
+            default_headers=default_headers or {},
+        )
+        self.openai = OpenAIClient(config)
+        self.anthropic = AnthropicClient(config)
+
+
+__all__ = [
+    "AllToken",
+    "AllTokenConfig",
+    "AllTokenError",
+    "AllTokenStream",
+    "AnthropicClient",
+    "ChatCompletion",
+    "ChatCompletionChunk",
+    "ChatCompletionMessage",
+    "OpenAIClient",
+    "__version__",
+]

alltoken-0.2.0/src/alltoken/_anthropic.py

@@ -0,0 +1,46 @@
+"""Anthropic-compatible client.
+
+Backed by ``anthropic.yml``. Base URL: ``https://api.alltoken.ai/anthropic``.
+"""
+from __future__ import annotations
+
+import httpx
+
+from ._client import AllTokenConfig, build_headers, join_base_url
+
+_ANTHROPIC_PATH = "/anthropic"
+
+
+class AnthropicClient:
+    """Anthropic-compatible client.
+
+    The underlying pre-configured ``httpx.Client`` is exposed as ``.raw``:
+
+        >>> resp = client.anthropic.raw.post(
+        ...     "/messages",
+        ...     json={
+        ...         "model": "claude-sonnet-4",
+        ...         "max_tokens": 1024,
+        ...         "messages": [{"role": "user", "content": "Hi"}],
+        ...     },
+        ... )
+    """
+
+    raw: httpx.Client
+
+    def __init__(self, config: AllTokenConfig) -> None:
+        self.raw = httpx.Client(
+            base_url=join_base_url(config.base_url, _ANTHROPIC_PATH),
+            headers=build_headers(config),
+            timeout=httpx.Timeout(60.0, connect=10.0),
+        )
+
+    def close(self) -> None:
+        """Close the underlying HTTP connection pool."""
+        self.raw.close()
+
+    def __enter__(self) -> AnthropicClient:
+        return self
+
+    def __exit__(self, *args: object) -> None:
+        self.close()

alltoken-0.2.0/src/alltoken/_client.py

@@ -0,0 +1,36 @@
+"""Shared base configuration for the OpenAI and Anthropic sub-clients.
+
+Both sub-clients share the same API key and base URL; only the path suffix
+differs (/v1 vs /anthropic).
+"""
+from __future__ import annotations
+
+from dataclasses import dataclass, field
+
+
+@dataclass
+class AllTokenConfig:
+    """Configuration for AllToken sub-clients."""
+
+    api_key: str
+    """API key from alltoken.ai. Found in Settings → API Keys."""
+
+    base_url: str = "https://api.alltoken.ai"
+    """Override the API base URL. Each sub-client appends its own path."""
+
+    default_headers: dict[str, str] = field(default_factory=dict)
+    """Extra headers sent on every request."""
+
+
+def build_headers(config: AllTokenConfig) -> dict[str, str]:
+    """Build the auth + default headers for every request."""
+    return {
+        "Authorization": f"Bearer {config.api_key}",
+        "Content-Type": "application/json",
+        **config.default_headers,
+    }
+
+
+def join_base_url(base_url: str, path: str) -> str:
+    """Join base URL + path, normalising trailing slashes."""
+    return base_url.rstrip("/") + path

alltoken-0.2.0/src/alltoken/_error.py

@@ -0,0 +1,9 @@
+"""AllToken API error type."""
+from __future__ import annotations
+
+
+class AllTokenError(Exception):
+    def __init__(self, status_code: int, body: str) -> None:
+        self.status_code = status_code
+        self.body = body
+        super().__init__(f"AllToken API error {status_code}: {body}")

alltoken-0.2.0/src/alltoken/_openai.py

@@ -0,0 +1,45 @@
+"""OpenAI-compatible client.
+
+Backed by ``chat.yml``. Base URL: ``https://api.alltoken.ai/v1``.
+"""
+from __future__ import annotations
+
+import httpx
+
+from ._client import AllTokenConfig, build_headers, join_base_url
+from ._resources.chat import Chat
+
+_OPENAI_PATH = "/v1"
+
+
+class OpenAIClient:
+    """OpenAI-compatible client.
+
+    The underlying pre-configured ``httpx.Client`` is exposed as ``.raw``:
+
+        >>> resp = client.openai.raw.post(
+        ...     "/chat/completions",
+        ...     json={"model": "gpt-4o", "messages": [{"role": "user", "content": "Hi"}]},
+        ... )
+    """
+
+    raw: httpx.Client
+    chat: Chat
+
+    def __init__(self, config: AllTokenConfig) -> None:
+        self.raw = httpx.Client(
+            base_url=join_base_url(config.base_url, _OPENAI_PATH),
+            headers=build_headers(config),
+            timeout=httpx.Timeout(60.0, connect=10.0),
+        )
+        self.chat = Chat(self.raw)
+
+    def close(self) -> None:
+        """Close the underlying HTTP connection pool."""
+        self.raw.close()
+
+    def __enter__(self) -> OpenAIClient:
+        return self
+
+    def __exit__(self, *args: object) -> None:
+        self.close()

alltoken-0.2.0/src/alltoken/_resources/chat.py

@@ -0,0 +1,50 @@
+"""Chat completions resource with typed convenience methods."""
+from __future__ import annotations
+
+from typing import Any, Literal, overload
+
+import httpx
+
+from .._error import AllTokenError
+from .._streaming import AllTokenStream
+from .._types import ChatCompletion, ChatCompletionChunk
+
+
+class ChatCompletions:
+    def __init__(self, client: httpx.Client) -> None:
+        self._client = client
+
+    @overload
+    def create(
+        self, *, stream: Literal[True], **kwargs: Any
+    ) -> AllTokenStream[ChatCompletionChunk]: ...
+
+    @overload
+    def create(
+        self, *, stream: Literal[False] = ..., **kwargs: Any
+    ) -> ChatCompletion: ...
+
+    def create(
+        self, *, stream: bool = False, **kwargs: Any
+    ) -> ChatCompletion | AllTokenStream[ChatCompletionChunk]:
+        body = {**kwargs, "stream": stream}
+        if stream:
+            request = self._client.build_request(
+                "POST", "/chat/completions", json=body
+            )
+            response = self._client.send(request, stream=True)
+            if response.status_code != 200:
+                body_text = response.read().decode()
+                response.close()
+                raise AllTokenError(response.status_code, body_text)
+            return AllTokenStream(response, ChatCompletionChunk)
+        else:
+            response = self._client.post("/chat/completions", json=body)
+            if response.status_code != 200:
+                raise AllTokenError(response.status_code, response.text)
+            return ChatCompletion.from_dict(response.json())
+
+
+class Chat:
+    def __init__(self, client: httpx.Client) -> None:
+        self.completions = ChatCompletions(client)

alltoken-0.2.0/src/alltoken/_streaming.py

@@ -0,0 +1,42 @@
+"""SSE stream iterator for AllToken streaming responses."""
+from __future__ import annotations
+
+import json
+from collections.abc import Iterator
+from typing import Any, Generic, Protocol, TypeVar
+
+import httpx
+
+T = TypeVar("T")
+
+
+class _FromDict(Protocol):
+    @classmethod
+    def from_dict(cls, data: dict[str, Any]) -> Any: ...
+
+
+class AllTokenStream(Generic[T]):
+    """Wraps an httpx streaming response, parses SSE lines into typed objects."""
+
+    def __init__(self, response: httpx.Response, parser: _FromDict) -> None:
+        self._response = response
+        self._parser = parser
+
+    def __iter__(self) -> Iterator[T]:
+        for line in self._response.iter_lines():
+            if not line.startswith("data: "):
+                continue
+            data = line[len("data: "):]
+            if data.strip() == "[DONE]":
+                break
+            obj = json.loads(data)
+            yield self._parser.from_dict(obj)
+
+    def close(self) -> None:
+        self._response.close()
+
+    def __enter__(self) -> AllTokenStream[T]:
+        return self
+
+    def __exit__(self, *args: object) -> None:
+        self.close()

alltoken-0.2.0/src/alltoken/_types.py

@@ -0,0 +1,102 @@
+"""Convenience dataclass types for chat completions."""
+from __future__ import annotations
+
+from dataclasses import dataclass
+from typing import Any
+
+
+@dataclass
+class ChatCompletionMessage:
+    role: str
+    content: str | None = None
+    tool_calls: list[Any] | None = None
+
+
+@dataclass
+class ChatCompletionChoice:
+    index: int
+    message: ChatCompletionMessage
+    finish_reason: str | None = None
+
+
+@dataclass
+class ChatCompletionUsage:
+    prompt_tokens: int
+    completion_tokens: int
+    total_tokens: int
+
+
+@dataclass
+class ChatCompletion:
+    id: str
+    object: str
+    created: int
+    model: str
+    choices: list[ChatCompletionChoice]
+    usage: ChatCompletionUsage | None = None
+
+    @classmethod
+    def from_dict(cls, data: dict[str, Any]) -> ChatCompletion:
+        choices = [
+            ChatCompletionChoice(
+                index=c["index"],
+                message=ChatCompletionMessage(**c["message"]),
+                finish_reason=c.get("finish_reason"),
+            )
+            for c in data["choices"]
+        ]
+        usage = ChatCompletionUsage(**data["usage"]) if data.get("usage") else None
+        return cls(
+            id=data["id"],
+            object=data["object"],
+            created=data["created"],
+            model=data["model"],
+            choices=choices,
+            usage=usage,
+        )
+
+
+# Streaming chunk types
+
+
+@dataclass
+class ChatCompletionChunkDelta:
+    role: str | None = None
+    content: str | None = None
+
+
+@dataclass
+class ChatCompletionChunkChoice:
+    index: int
+    delta: ChatCompletionChunkDelta
+    finish_reason: str | None = None
+
+
+@dataclass
+class ChatCompletionChunk:
+    id: str
+    object: str
+    created: int
+    model: str
+    choices: list[ChatCompletionChunkChoice]
+    usage: ChatCompletionUsage | None = None
+
+    @classmethod
+    def from_dict(cls, data: dict[str, Any]) -> ChatCompletionChunk:
+        choices = [
+            ChatCompletionChunkChoice(
+                index=c["index"],
+                delta=ChatCompletionChunkDelta(**c.get("delta", {})),
+                finish_reason=c.get("finish_reason"),
+            )
+            for c in data["choices"]
+        ]
+        usage = ChatCompletionUsage(**data["usage"]) if data.get("usage") else None
+        return cls(
+            id=data["id"],
+            object=data["object"],
+            created=data["created"],
+            model=data["model"],
+            choices=choices,
+            usage=usage,
+        )

alltoken-0.2.0/src/alltoken/generated/__init__.py

@@ -0,0 +1,6 @@
+"""Pydantic models generated from the AllToken OpenAPI specs.
+
+Run ``python scripts/generate.py`` from the repo root to regenerate after the
+specs move. These modules are committed so users who install from PyPI don't
+need to run codegen.
+"""

alltoken-0.2.0/src/alltoken/generated/anthropic.py

@@ -0,0 +1,201 @@
+# generated by datamodel-codegen:
+#   filename:  anthropic.yml
+#   timestamp: 2026-04-15T09:50:44+00:00
+
+from __future__ import annotations
+
+from enum import Enum
+from typing import Annotated, Any
+
+from pydantic import AwareDatetime, BaseModel, Field, RootModel
+
+
+class Metadata(BaseModel):
+    user_id: str | None = None
+    """
+    终端用户标识
+    """
+
+
+class Type(Enum):
+    enabled = "enabled"
+
+
+class Thinking(BaseModel):
+    """
+    Extended thinking 配置
+    """
+
+    type: Type | None = None
+    budget_tokens: int | None = None
+
+
+class Role(Enum):
+    user = "user"
+    assistant = "assistant"
+
+
+class Type1(Enum):
+    text = "text"
+    image = "image"
+    tool_use = "tool_use"
+    tool_result = "tool_result"
+
+
+class ContentBlock(BaseModel):
+    type: Type1
+    text: str | None = None
+    """
+    type=text 时的文本内容
+    """
+    id: str | None = None
+    """
+    type=tool_use 时的工具调用 ID
+    """
+    name: str | None = None
+    """
+    type=tool_use 时的工具名
+    """
+    input: Any | None = None
+    """
+    type=tool_use 时的工具参数（任意 JSON）
+    """
+    cache_control: dict[str, Any] | None = None
+    """
+    块级缓存控制，如 `{type: ephemeral, ttl: 1h}`
+    """
+
+
+class Tool(BaseModel):
+    name: str
+    description: str | None = None
+    input_schema: dict[str, Any]
+    """
+    JSON Schema 对象
+    """
+    cache_control: dict[str, Any] | None = None
+
+
+class Type2(Enum):
+    message = "message"
+
+
+class Role1(Enum):
+    assistant = "assistant"
+
+
+class StopReasonEnum(Enum):
+    end_turn = "end_turn"
+    max_tokens = "max_tokens"
+    stop_sequence = "stop_sequence"
+    tool_use = "tool_use"
+
+
+class StopReason(RootModel[StopReasonEnum | None]):
+    root: StopReasonEnum | None = None
+
+
+class Usage(BaseModel):
+    input_tokens: int
+    output_tokens: int
+    cache_creation_input_tokens: int | None = None
+    cache_read_input_tokens: int | None = None
+
+
+class Type3(Enum):
+    model = "model"
+
+
+class ModelInfo(BaseModel):
+    id: str
+    type: Type3
+    display_name: str | None = None
+    created_at: AwareDatetime | None = None
+
+
+class Type4(Enum):
+    error = "error"
+
+
+class Type5(Enum):
+    invalid_request_error = "invalid_request_error"
+    authentication_error = "authentication_error"
+    permission_error = "permission_error"
+    not_found_error = "not_found_error"
+    rate_limit_error = "rate_limit_error"
+    api_error = "api_error"
+    overloaded_error = "overloaded_error"
+
+
+class Error(BaseModel):
+    type: Type5
+    message: str
+
+
+class ErrorResponse(BaseModel):
+    type: Type4
+    error: Error
+
+
+class InputMessage(BaseModel):
+    role: Role
+    content: str | list[ContentBlock]
+    """
+    字符串或 ContentBlock 数组
+    """
+
+
+class MessageResponse(BaseModel):
+    id: Annotated[str, Field(examples=["msg_01XFDUDYJgAACzvnptvVoYEL"])]
+    type: Type2
+    role: Role1
+    content: list[ContentBlock]
+    model: str
+    stop_reason: StopReason
+    stop_sequence: str | None = None
+    usage: Usage
+
+
+class ModelList(BaseModel):
+    data: list[ModelInfo]
+    has_more: bool | None = None
+    first_id: str | None = None
+    last_id: str | None = None
+
+
+class MessageRequest(BaseModel):
+    model: Annotated[str, Field(examples=["claude-sonnet-4-5"])]
+    """
+    模型 ID（见 GET /v1/models）
+    """
+    max_tokens: Annotated[int, Field(examples=[1024], ge=1)]
+    """
+    生成的最大 token 数
+    """
+    messages: list[InputMessage]
+    system: str | list[ContentBlock] | None = None
+    """
+    System prompt（字符串 或 ContentBlock 数组）
+    """
+    stream: bool | None = False
+    """
+    是否以 SSE 流式返回
+    """
+    temperature: Annotated[float | None, Field(ge=0.0, le=1.0)] = None
+    top_p: Annotated[float | None, Field(ge=0.0, le=1.0)] = None
+    top_k: Annotated[int | None, Field(ge=0)] = None
+    stop_sequences: list[str] | None = None
+    tools: list[Tool] | None = None
+    tool_choice: Any | None = None
+    """
+    工具选择策略：`{type: auto}` / `{type: any}` / `{type: tool, name: ...}`
+    """
+    metadata: Metadata | None = None
+    thinking: Thinking | None = None
+    """
+    Extended thinking 配置
+    """
+    cache_control: dict[str, Any] | None = None
+    """
+    顶层自动缓存控制，如 `{type: ephemeral, ttl: 1h}`
+    """

alltoken-0.2.0/src/alltoken/generated/chat.py

@@ -0,0 +1,426 @@
+# generated by datamodel-codegen:
+#   filename:  chat.yml
+#   timestamp: 2026-04-15T09:50:43+00:00
+
+from __future__ import annotations
+
+from enum import Enum
+from typing import Annotated, Any
+
+from pydantic import AnyUrl, AwareDatetime, BaseModel, Field, RootModel
+
+
+class StreamOptions(BaseModel):
+    include_usage: bool | None = None
+
+
+class Type(Enum):
+    text = "text"
+    json_object = "json_object"
+
+
+class ResponseFormat(BaseModel):
+    type: Type | None = None
+
+
+class Role(Enum):
+    system = "system"
+    user = "user"
+    assistant = "assistant"
+    tool = "tool"
+
+
+class Type1(Enum):
+    text = "text"
+    image_url = "image_url"
+
+
+class Detail(Enum):
+    auto = "auto"
+    low = "low"
+    high = "high"
+
+
+class ImageUrl(BaseModel):
+    """
+    type=image_url 时的图片对象
+    """
+
+    url: AnyUrl | None = None
+    detail: Detail | None = None
+
+
+class ContentPart(BaseModel):
+    """
+    多模态消息内容块（文本 / 图片 URL）
+    """
+
+    type: Type1
+    text: str | None = None
+    """
+    type=text 时的文本内容
+    """
+    image_url: ImageUrl | None = None
+    """
+    type=image_url 时的图片对象
+    """
+
+
+class Type2(Enum):
+    function = "function"
+
+
+class Function(BaseModel):
+    name: str
+    description: str | None = None
+    parameters: dict[str, Any] | None = None
+    """
+    JSON Schema
+    """
+
+
+class Tool(BaseModel):
+    type: Type2
+    function: Function
+
+
+class Function1(BaseModel):
+    name: str | None = None
+    arguments: str | None = None
+    """
+    JSON 字符串
+    """
+
+
+class ToolCall(BaseModel):
+    id: str
+    type: Type2
+    function: Function1
+
+
+class Object(Enum):
+    chat_completion = "chat.completion"
+
+
+class FinishReasonEnum(Enum):
+    stop = "stop"
+    length = "length"
+    tool_calls = "tool_calls"
+    content_filter = "content_filter"
+
+
+class FinishReason(RootModel[FinishReasonEnum | None]):
+    root: FinishReasonEnum | None = None
+
+
+class PromptTokensDetails(BaseModel):
+    cached_tokens: int | None = None
+    cache_creation_input_tokens: int | None = None
+
+
+class CompletionTokensDetails(BaseModel):
+    reasoning_tokens: int | None = None
+
+
+class Usage(BaseModel):
+    prompt_tokens: int
+    completion_tokens: int
+    total_tokens: int
+    prompt_tokens_details: PromptTokensDetails | None = None
+    completion_tokens_details: CompletionTokensDetails | None = None
+    cache_read_input_tokens: int | None = None
+    cache_creation_by_ttl: dict[str, int] | None = None
+    """
+    Cache write 按 TTL 拆分，key=秒数（如 300、3600）
+    """
+    server_tool_use: dict[str, int] | None = None
+
+
+class Object1(Enum):
+    list = "list"
+
+
+class Object2(Enum):
+    model = "model"
+
+
+class ModelInfo(BaseModel):
+    id: str
+    object: Object2
+    created: int | None = None
+    owned_by: str | None = None
+
+
+class Ratio(Enum):
+    field_16_9 = "16:9"
+    field_9_16 = "9:16"
+    field_4_3 = "4:3"
+    field_3_4 = "3:4"
+    field_21_9 = "21:9"
+    field_1_1 = "1:1"
+    adaptive = "adaptive"
+
+
+class Resolution(Enum):
+    field_480p = "480p"
+    field_720p = "720p"
+    field_1080p = "1080p"
+
+
+class ServiceTier(Enum):
+    default = "default"
+    flex = "flex"
+
+
+class Type4(Enum):
+    web_search = "web_search"
+
+
+class Tool1(BaseModel):
+    type: Type4 | None = None
+
+
+class Type5(Enum):
+    text = "text"
+    image_url = "image_url"
+    video_url = "video_url"
+    audio_url = "audio_url"
+    draft_task = "draft_task"
+    image = "image"
+    video = "video"
+    audio = "audio"
+
+
+class ImageUrl1(BaseModel):
+    url: AnyUrl | None = None
+
+
+class VideoUrl(BaseModel):
+    url: AnyUrl | None = None
+
+
+class AudioUrl(BaseModel):
+    url: AnyUrl | None = None
+
+
+class DraftTask(BaseModel):
+    id: str | None = None
+
+
+class Role1(Enum):
+    first_frame = "first_frame"
+    last_frame = "last_frame"
+    reference_image = "reference_image"
+    reference_video = "reference_video"
+    reference_audio = "reference_audio"
+
+
+class VideoContentItem(BaseModel):
+    type: Type5
+    text: str | None = None
+    image_url: ImageUrl1 | None = None
+    video_url: VideoUrl | None = None
+    audio_url: AudioUrl | None = None
+    draft_task: DraftTask | None = None
+    role: Role1 | None = None
+    url: str | None = None
+    """
+    简化格式（type=image/video/audio 时）
+    """
+
+
+class Status(Enum):
+    queued = "queued"
+    processing = "processing"
+    completed = "completed"
+    failed = "failed"
+    expired = "expired"
+    cancelled = "cancelled"
+
+
+class InputType(Enum):
+    text = "text"
+    image_first_frame = "image_first_frame"
+    image_first_last_frame = "image_first_last_frame"
+    image_reference = "image_reference"
+    video_reference = "video_reference"
+    multimodal = "multimodal"
+    draft_task = "draft_task"
+
+
+class ToolUsage(BaseModel):
+    web_search: int | None = None
+
+
+class Usage1(BaseModel):
+    completion_tokens: int | None = None
+    total_tokens: int | None = None
+    tool_usage: ToolUsage | None = None
+
+
+class Error(BaseModel):
+    code: str | None = None
+    message: str | None = None
+
+
+class VideoTaskResponse(BaseModel):
+    id: Annotated[str, Field(examples=["vgen_01abc"])]
+    status: Status
+    model: str
+    input_type: InputType | None = None
+    video_url: AnyUrl | None = None
+    video_url_expires_at: AwareDatetime | None = None
+    video_url_ttl: int | None = None
+    """
+    视频 URL 剩余有效秒数
+    """
+    last_frame_url: AnyUrl | None = None
+    duration: int | None = None
+    frames: int | None = None
+    fps: int | None = None
+    resolution: str | None = None
+    ratio: str | None = None
+    seed: int | None = None
+    generate_audio: bool | None = None
+    draft: bool | None = None
+    draft_task_id: str | None = None
+    service_tier: str | None = None
+    execution_expires_after: int | None = None
+    safety_identifier: str | None = None
+    tools: list[dict[str, Any]] | None = None
+    usage: Usage1 | None = None
+    error: Error | None = None
+    created_at: AwareDatetime
+    updated_at: AwareDatetime | None = None
+    completed_at: AwareDatetime | None = None
+
+
+class VideoListResponse(BaseModel):
+    items: list[VideoTaskResponse]
+    total: int
+
+
+class Error1(BaseModel):
+    code: str | None = None
+    message: str
+    param: str | None = None
+    type: str
+    request_id: str | None = None
+
+
+class ErrorResponse(BaseModel):
+    error: Error1
+
+
+class ChatMessage(BaseModel):
+    role: Role
+    content: str | list[ContentPart] | None = None
+    """
+    消息内容：字符串，或 ContentPart 数组（多模态），或 null（assistant 返回 tool_calls 时）。
+    JSON 层允许 null；SDK 用 Union/any 类型接收。
+
+    """
+    name: str | None = None
+    tool_calls: list[ToolCall] | None = None
+    tool_call_id: str | None = None
+    """
+    role=tool 时必填
+    """
+
+
+class Choice(BaseModel):
+    index: int
+    message: ChatMessage
+    finish_reason: FinishReason | None = None
+
+
+class ModelList(BaseModel):
+    object: Object1
+    data: list[ModelInfo]
+
+
+class VideoGenerationRequest(BaseModel):
+    model: Annotated[str, Field(examples=["seedance-1.5-pro"])]
+    prompt: str
+    content: list[VideoContentItem] | None = None
+    ratio: Ratio | None = None
+    duration: int | None = None
+    """
+    秒数；-1 表示智能时长
+    """
+    resolution: Resolution | None = None
+    frames: int | None = None
+    """
+    帧数（优先级高于 duration）
+    """
+    generate_audio: bool | None = True
+    seed: int | None = None
+    camera_fixed: bool | None = None
+    watermark: bool | None = False
+    callback_url: AnyUrl | None = None
+    return_last_frame: bool | None = None
+    service_tier: ServiceTier | None = None
+    execution_expires_after: Annotated[int | None, Field(ge=3600, le=259200)] = None
+    draft: bool | None = None
+    tools: list[Tool1] | None = None
+    safety_identifier: Annotated[str | None, Field(max_length=64)] = None
+
+
+class ChatCompletionRequest(BaseModel):
+    model: Annotated[str, Field(examples=["gpt-4o-mini"])]
+    messages: list[ChatMessage]
+    stream: bool | None = False
+    stream_options: StreamOptions | None = None
+    max_tokens: Annotated[int | None, Field(ge=1)] = None
+    temperature: Annotated[float | None, Field(ge=0.0, le=2.0)] = None
+    top_p: float | None = None
+    top_k: int | None = None
+    stop: list[str] | None = None
+    user: str | None = None
+    tools: list[Tool] | None = None
+    tool_choice: Any | None = None
+    """
+    `auto` / `none` / `required` / `{type: function, function: {name: ...}}`
+    """
+    response_format: ResponseFormat | None = None
+    enable_search: bool | None = None
+    """
+    网关扩展 —— 启用联网搜索
+    """
+    frequency_penalty: float | None = None
+    presence_penalty: float | None = None
+    n: Annotated[int | None, Field(ge=1)] = None
+    thinking: bool | None = None
+    """
+    网关扩展 —— 启用思考链
+    """
+    thinking_budget: int | None = None
+    """
+    网关扩展 —— 思考链 token 预算
+    """
+    conversation_id: int | None = None
+    """
+    网关扩展 —— 关联网关托管的会话 ID（自动持久化消息）
+    """
+
+
+class ChatCompletionResponse(BaseModel):
+    id: str
+    object: Object
+    created: int
+    model: str
+    choices: list[Choice]
+    usage: Usage | None = None
+    system_fingerprint: str | None = None
+    x_gateway_provider: Annotated[str | None, Field(alias="x-gateway-provider")] = None
+    """
+    网关扩展 —— 实际处理请求的 provider 名
+    """
+    x_gateway_request_id: Annotated[str | None, Field(alias="x-gateway-request-id")] = None
+    """
+    网关扩展 —— 请求 ID（便于排查）
+    """
+    conversation_id: int | None = None
+    """
+    网关扩展 —— 若请求触发自动建会话，返回新会话 ID
+    """