runapi-seedream 0.1.0__tar.gz

runapi_seedream-0.1.0/PKG-INFO

@@ -0,0 +1,78 @@
+Metadata-Version: 2.4
+Name: runapi-seedream
+Version: 0.1.0
+Summary: Seedream text-to-image and edit-image client for RunAPI
+Project-URL: Homepage, https://runapi.ai/models/seedream
+Project-URL: Documentation, https://runapi.ai/docs#sdk-seedream
+Author-email: RunAPI <contact@runapi.ai>
+License-Expression: Apache-2.0
+Keywords: ai,edit-image,runapi,sdk,seedream,text-to-image
+Requires-Python: >=3.9
+Requires-Dist: runapi-core
+Description-Content-Type: text/markdown
+
+# Seedream API Python SDK for RunAPI
+
+The seedream api Python SDK is the language-specific package for Seedream on RunAPI. Use this seedream api package for text-to-image, image editing, and creative production flows when your application needs JSON request bodies, task status lookup, and consistent RunAPI errors in Python.
+
+This seedream api README is the Python package guide inside the public `seedream-sdk` repository. For the repository overview, start at `../README.md`; for model details, use https://runapi.ai/models/seedream; for API reference, use https://runapi.ai/docs#seedream; for SDK docs, use https://runapi.ai/docs#sdk-seedream.
+
+## Install
+
+```bash
+pip install runapi-seedream
+```
+
+## Quick start
+
+```python
+from runapi.seedream import SeedreamClient
+
+client = SeedreamClient()  # reads RUNAPI_API_KEY, or pass api_key="sk-..."
+
+task = client.text_to_image.create(
+    model="seedream-v4-text-to-image",
+    prompt="A precise product render of a glass teapot on white marble",
+    aspect_ratio="16:9",
+    output_resolution="2k",
+    output_count=3,
+)
+status = client.text_to_image.get(task.id)
+
+edit = client.edit_image.create(
+    model="seedream-v4-edit",
+    prompt="Make it golden hour",
+    source_image_urls=["https://example.com/source.jpg"],
+)
+```
+
+Use `create` to submit a task and return quickly, `get` to fetch the latest task state, and `run` when a script should create and poll until completion:
+
+```python
+result = client.text_to_image.run(
+    model="seedream-v4-text-to-image",
+    prompt="A serene mountain lake at dawn",
+)
+print(result.images[0].url)
+```
+
+In web request handlers, prefer `create` plus webhook or later `get` polling so a worker is not held open.
+
+RunAPI-generated file URLs are temporary. Download and store generated images, videos, audio, or other files in your own durable storage within 7 days; do not treat returned URLs as long-term assets.
+
+## Language notes
+
+Pass parameters as keyword arguments and catch the `runapi.seedream` error classes when building image jobs or scripts. The available resources are `text_to_image` for text models and `edit_image` for editing models. Keep `RUNAPI_API_KEY` in the environment or your secret manager; never commit API keys or callback secrets.
+
+## Links
+
+- Model page: https://runapi.ai/models/seedream
+- SDK docs: https://runapi.ai/docs#sdk-seedream
+- Product docs: https://runapi.ai/docs#seedream
+- Pricing and rate limits: https://runapi.ai/models/seedream/v4-text-to-image
+- Full catalog: https://runapi.ai/models
+- Repository: https://github.com/runapi-ai/seedream-sdk
+
+## License
+
+Licensed under the Apache License, Version 2.0.

runapi_seedream-0.1.0/README.md

@@ -0,0 +1,65 @@
+# Seedream API Python SDK for RunAPI
+
+The seedream api Python SDK is the language-specific package for Seedream on RunAPI. Use this seedream api package for text-to-image, image editing, and creative production flows when your application needs JSON request bodies, task status lookup, and consistent RunAPI errors in Python.
+
+This seedream api README is the Python package guide inside the public `seedream-sdk` repository. For the repository overview, start at `../README.md`; for model details, use https://runapi.ai/models/seedream; for API reference, use https://runapi.ai/docs#seedream; for SDK docs, use https://runapi.ai/docs#sdk-seedream.
+
+## Install
+
+```bash
+pip install runapi-seedream
+```
+
+## Quick start
+
+```python
+from runapi.seedream import SeedreamClient
+
+client = SeedreamClient()  # reads RUNAPI_API_KEY, or pass api_key="sk-..."
+
+task = client.text_to_image.create(
+    model="seedream-v4-text-to-image",
+    prompt="A precise product render of a glass teapot on white marble",
+    aspect_ratio="16:9",
+    output_resolution="2k",
+    output_count=3,
+)
+status = client.text_to_image.get(task.id)
+
+edit = client.edit_image.create(
+    model="seedream-v4-edit",
+    prompt="Make it golden hour",
+    source_image_urls=["https://example.com/source.jpg"],
+)
+```
+
+Use `create` to submit a task and return quickly, `get` to fetch the latest task state, and `run` when a script should create and poll until completion:
+
+```python
+result = client.text_to_image.run(
+    model="seedream-v4-text-to-image",
+    prompt="A serene mountain lake at dawn",
+)
+print(result.images[0].url)
+```
+
+In web request handlers, prefer `create` plus webhook or later `get` polling so a worker is not held open.
+
+RunAPI-generated file URLs are temporary. Download and store generated images, videos, audio, or other files in your own durable storage within 7 days; do not treat returned URLs as long-term assets.
+
+## Language notes
+
+Pass parameters as keyword arguments and catch the `runapi.seedream` error classes when building image jobs or scripts. The available resources are `text_to_image` for text models and `edit_image` for editing models. Keep `RUNAPI_API_KEY` in the environment or your secret manager; never commit API keys or callback secrets.
+
+## Links
+
+- Model page: https://runapi.ai/models/seedream
+- SDK docs: https://runapi.ai/docs#sdk-seedream
+- Product docs: https://runapi.ai/docs#seedream
+- Pricing and rate limits: https://runapi.ai/models/seedream/v4-text-to-image
+- Full catalog: https://runapi.ai/models
+- Repository: https://github.com/runapi-ai/seedream-sdk
+
+## License
+
+Licensed under the Apache License, Version 2.0.

runapi_seedream-0.1.0/pyproject.toml

@@ -0,0 +1,33 @@
+[build-system]
+requires = ["hatchling"]
+build-backend = "hatchling.build"
+
+[project]
+name = "runapi-seedream"
+version = "0.1.0"
+description = "Seedream text-to-image and edit-image client for RunAPI"
+readme = "README.md"
+requires-python = ">=3.9"
+license = "Apache-2.0"
+authors = [{ name = "RunAPI", email = "contact@runapi.ai" }]
+keywords = ["runapi", "seedream", "text-to-image", "edit-image", "ai", "sdk"]
+dependencies = ["runapi-core"]
+
+[project.urls]
+Homepage = "https://runapi.ai/models/seedream"
+Documentation = "https://runapi.ai/docs#sdk-seedream"
+
+[tool.hatch.build.targets.wheel]
+packages = ["src/runapi"]
+
+[tool.uv]
+package = true
+
+[tool.uv.sources]
+runapi-core = { workspace = true }
+
+[dependency-groups]
+dev = ["pytest>=8"]
+
+[tool.runapi]
+slug = "seedream"

runapi_seedream-0.1.0/src/runapi/seedream/__init__.py

@@ -0,0 +1,24 @@
+"""Seedream client for RunAPI."""
+
+from runapi.core import (
+    AuthenticationError,
+    InsufficientCreditsError,
+    NotFoundError,
+    RateLimitError,
+    TaskFailedError,
+    TaskTimeoutError,
+    ValidationError,
+)
+
+from .client import SeedreamClient
+
+__all__ = [
+    "SeedreamClient",
+    "AuthenticationError",
+    "RateLimitError",
+    "InsufficientCreditsError",
+    "NotFoundError",
+    "ValidationError",
+    "TaskFailedError",
+    "TaskTimeoutError",
+]

runapi_seedream-0.1.0/src/runapi/seedream/client.py

@@ -0,0 +1,29 @@
+"""Seedream client."""
+
+from __future__ import annotations
+
+from typing import Any, Optional
+
+from runapi.core import ClientOptions, HttpClient, resolve_api_key
+
+from .resources.edit_image import EditImage
+from .resources.text_to_image import TextToImage
+
+
+class SeedreamClient:
+    """Seedream text-to-image and edit-image client.
+
+    Example::
+
+        client = SeedreamClient(api_key="sk-...")
+        result = client.text_to_image.run(
+            model="seedream-v4-text-to-image", prompt="A neon city street"
+        )
+    """
+
+    def __init__(self, api_key: Optional[str] = None, **options: Any) -> None:
+        resolved_api_key = resolve_api_key(api_key)
+        client_options = ClientOptions(api_key=resolved_api_key, **options)
+        http = client_options.http_client or HttpClient(client_options)
+        self.text_to_image = TextToImage(http)
+        self.edit_image = EditImage(http)

runapi_seedream-0.1.0/src/runapi/seedream/contract_gen.py

@@ -0,0 +1,117 @@
+# Code generated by SdkContractGenerator. DO NOT EDIT.
+
+CONTRACT = {
+    "edit-image": {
+        "models": ["seedream-4.5-edit", "seedream-5-lite-edit", "seedream-v4-edit"],
+        "fields_by_model": {
+            "seedream-4.5-edit": {
+                "aspect_ratio": {
+                    "enum": ["1:1", "4:3", "3:4", "16:9", "9:16", "2:3", "3:2", "21:9"],
+                    "required": True
+                },
+                "output_count": {
+                    "type": "integer"
+                },
+                "output_quality": {
+                    "enum": ["basic", "high"],
+                    "required": True
+                },
+                "seed": {
+                    "type": "integer"
+                },
+                "source_image_urls": {
+                    "required": True
+                }
+            },
+            "seedream-5-lite-edit": {
+                "aspect_ratio": {
+                    "enum": ["1:1", "4:3", "3:4", "16:9", "9:16", "2:3", "3:2", "21:9"],
+                    "required": True
+                },
+                "output_count": {
+                    "type": "integer"
+                },
+                "output_quality": {
+                    "enum": ["basic", "high"],
+                    "required": True
+                },
+                "seed": {
+                    "type": "integer"
+                },
+                "source_image_urls": {
+                    "required": True
+                }
+            },
+            "seedream-v4-edit": {
+                "aspect_ratio": {
+                    "enum": ["1:1", "4:3", "3:4", "3:2", "2:3", "16:9", "9:16", "21:9"]
+                },
+                "output_count": {
+                    "enum": [1, 2, 3, 4, 5, 6],
+                    "type": "integer"
+                },
+                "output_resolution": {
+                    "enum": ["1k", "2k", "4k"]
+                },
+                "seed": {
+                    "type": "integer"
+                },
+                "source_image_urls": {
+                    "required": True
+                }
+            }
+        }
+    },
+    "text-to-image": {
+        "models": ["seedream-4.5-text-to-image", "seedream-5-lite-text-to-image", "seedream-v4-text-to-image"],
+        "fields_by_model": {
+            "seedream-4.5-text-to-image": {
+                "aspect_ratio": {
+                    "enum": ["1:1", "4:3", "3:4", "16:9", "9:16", "2:3", "3:2", "21:9"],
+                    "required": True
+                },
+                "output_count": {
+                    "type": "integer"
+                },
+                "output_quality": {
+                    "enum": ["basic", "high"],
+                    "required": True
+                },
+                "seed": {
+                    "type": "integer"
+                }
+            },
+            "seedream-5-lite-text-to-image": {
+                "aspect_ratio": {
+                    "enum": ["1:1", "4:3", "3:4", "16:9", "9:16", "2:3", "3:2", "21:9"],
+                    "required": True
+                },
+                "output_count": {
+                    "type": "integer"
+                },
+                "output_quality": {
+                    "enum": ["basic", "high"],
+                    "required": True
+                },
+                "seed": {
+                    "type": "integer"
+                }
+            },
+            "seedream-v4-text-to-image": {
+                "aspect_ratio": {
+                    "enum": ["1:1", "4:3", "3:4", "3:2", "2:3", "16:9", "9:16", "21:9"]
+                },
+                "output_count": {
+                    "enum": [1, 2, 3, 4, 5, 6],
+                    "type": "integer"
+                },
+                "output_resolution": {
+                    "enum": ["1k", "2k", "4k"]
+                },
+                "seed": {
+                    "type": "integer"
+                }
+            }
+        }
+    }
+}

runapi_seedream-0.1.0/src/runapi/seedream/resources/__init__.py

@@ -0,0 +1,4 @@
+from .edit_image import EditImage
+from .text_to_image import TextToImage
+
+__all__ = ["TextToImage", "EditImage"]

runapi_seedream-0.1.0/src/runapi/seedream/resources/edit_image.py

@@ -0,0 +1,92 @@
+"""Seedream edit-image resource."""
+
+from __future__ import annotations
+
+from typing import Any, Dict
+
+from runapi.core import Resource, ValidationError
+
+from ..contract_gen import CONTRACT
+from ..types import (
+    LITE_MODELS,
+    V4_MODELS,
+    CompletedEditImageResponse,
+    EditImageResponse,
+)
+
+
+class EditImage(Resource):
+    """Edit images from a prompt and source images with Seedream models."""
+
+    ENDPOINT = "/api/v1/seedream/edit_image"
+
+    RESPONSE_CLASS = EditImageResponse
+    COMPLETED_RESPONSE_CLASS = CompletedEditImageResponse
+
+    PROMPT_MAX_LENGTH = 3000
+    V4_PROMPT_MAX_LENGTH = 5000
+    PROMPT_MIN_LENGTH_LITE = 3
+
+    def run(self, **params: Any) -> Any:
+        """Edit an image and poll until it completes.
+
+        Args:
+            **params: image edit parameters (model, ...).
+
+        Returns:
+            The completed (narrowed) image edit response.
+        """
+        task = self.create(**params)
+        return self._poll_until_complete(lambda: self.get(task.id))
+
+    def create(self, **params: Any) -> Any:
+        """Create an image editing task and return immediately with an id.
+
+        Args:
+            **params: image edit parameters (model, ...).
+
+        Returns:
+            The task creation result with an id.
+        """
+        compacted = self._compact_params(params)
+        self._validate_params(compacted)
+        return self._request("post", self.ENDPOINT, body=compacted)
+
+    def get(self, id: str) -> Any:
+        """Fetch the current status of an image editing task.
+
+        Args:
+            id: The task id returned by ``create``.
+
+        Returns:
+            The current task status.
+        """
+        return self._request("get", f"{self.ENDPOINT}/{id}")
+
+    def _validate_params(self, params: Dict[str, Any]) -> None:
+        self._validate_contract(CONTRACT["edit-image"], params)
+
+        model = params.get("model")
+
+        prompt = params.get("prompt")
+        if not (isinstance(prompt, str) and prompt):
+            raise ValidationError("prompt is required")
+        max_length = self.V4_PROMPT_MAX_LENGTH if model in V4_MODELS else self.PROMPT_MAX_LENGTH
+        if len(prompt) > max_length:
+            raise ValidationError(f"prompt must be at most {max_length} characters")
+        if model in LITE_MODELS and len(prompt) < self.PROMPT_MIN_LENGTH_LITE:
+            raise ValidationError(
+                f"prompt must be between {self.PROMPT_MIN_LENGTH_LITE} and {self.PROMPT_MAX_LENGTH} characters"
+            )
+
+        if model in V4_MODELS:
+            self._validate_integer(params, "seed")
+
+    @staticmethod
+    def _validate_integer(params: Dict[str, Any], key: str) -> None:
+        value = params.get(key)
+        if value is None:
+            return
+        if isinstance(value, int) and not isinstance(value, bool):
+            return
+        raise ValidationError(f"{key} must be an integer")

runapi_seedream-0.1.0/src/runapi/seedream/resources/text_to_image.py

@@ -0,0 +1,104 @@
+"""Seedream text-to-image resource."""
+
+from __future__ import annotations
+
+from typing import Any, Dict
+
+from runapi.core import Resource, ValidationError
+
+from ..contract_gen import CONTRACT
+from ..types import (
+    LITE_MODELS,
+    V4_MODELS,
+    CompletedTextToImageResponse,
+    TextToImageResponse,
+)
+
+
+class TextToImage(Resource):
+    """Generate images from text prompts with Seedream models."""
+
+    ENDPOINT = "/api/v1/seedream/text_to_image"
+
+    RESPONSE_CLASS = TextToImageResponse
+    COMPLETED_RESPONSE_CLASS = CompletedTextToImageResponse
+
+    PROMPT_MAX_LENGTH = 3000
+    V4_PROMPT_MAX_LENGTH = 5000
+    PROMPT_MIN_LENGTH_LITE = 3
+
+    def run(self, **params: Any) -> Any:
+        """Create a text-to-image task and poll until it completes.
+
+        Args:
+            **params: text-to-image parameters (model, ...).
+
+        Returns:
+            The completed (narrowed) text-to-image response.
+        """
+        task = self.create(**params)
+        return self._poll_until_complete(lambda: self.get(task.id))
+
+    def create(self, **params: Any) -> Any:
+        """Create a text-to-image task and return immediately with an id.
+
+        Args:
+            **params: text-to-image parameters (model, ...).
+
+        Returns:
+            The task creation result with an id.
+        """
+        compacted = self._compact_params(params)
+        self._validate_params(compacted)
+        return self._request("post", self.ENDPOINT, body=compacted)
+
+    def get(self, id: str) -> Any:
+        """Fetch the current status of a text-to-image task.
+
+        Args:
+            id: The task id returned by ``create``.
+
+        Returns:
+            The current task status.
+        """
+        return self._request("get", f"{self.ENDPOINT}/{id}")
+
+    def _validate_params(self, params: Dict[str, Any]) -> None:
+        self._validate_contract(CONTRACT["text-to-image"], params)
+
+        model = params.get("model")
+
+        prompt = params.get("prompt")
+        if not (isinstance(prompt, str) and prompt):
+            raise ValidationError("prompt is required")
+        max_length = self.V4_PROMPT_MAX_LENGTH if model in V4_MODELS else self.PROMPT_MAX_LENGTH
+        if len(prompt) > max_length:
+            raise ValidationError(f"prompt must be at most {max_length} characters")
+        if model in LITE_MODELS and len(prompt) < self.PROMPT_MIN_LENGTH_LITE:
+            raise ValidationError(
+                f"prompt must be between {self.PROMPT_MIN_LENGTH_LITE} and {self.PROMPT_MAX_LENGTH} characters"
+            )
+
+        if model in V4_MODELS:
+            self._validate_integer(params, "seed")
+
+        if self._field_present(params, "source_image_urls"):
+            raise ValidationError(f"source_image_urls is not supported for {model}")
+
+    @staticmethod
+    def _field_present(params: Dict[str, Any], key: str) -> bool:
+        value = params.get(key)
+        if value is None:
+            return False
+        if hasattr(value, "__len__"):
+            return len(value) > 0
+        return True
+
+    @staticmethod
+    def _validate_integer(params: Dict[str, Any], key: str) -> None:
+        value = params.get(key)
+        if value is None:
+            return
+        if isinstance(value, int) and not isinstance(value, bool):
+            return
+        raise ValidationError(f"{key} must be an integer")

runapi_seedream-0.1.0/src/runapi/seedream/types.py

@@ -0,0 +1,36 @@
+"""Seedream model lists, enums, and response models."""
+
+from __future__ import annotations
+
+from runapi.core import BaseModel, TaskResponse, optional, required
+
+# Model groupings used by bespoke prompt-length selection (per-model rule the
+# contract cannot express). Model membership and field enums are validated by
+# the generated CONTRACT.
+LITE_MODELS = ["seedream-5-lite-text-to-image", "seedream-5-lite-edit"]
+V4_MODELS = ["seedream-v4-text-to-image", "seedream-v4-edit"]
+
+
+class Image(BaseModel):
+    url = optional(str)
+
+
+class TextToImageResponse(TaskResponse):
+    """Seedream image task status response."""
+
+    id = required(str)
+    status = optional(str, enum=lambda: TaskResponse.Status.ALL)
+    images = optional([lambda: Image])
+    error = optional(str)
+
+
+EditImageResponse = TextToImageResponse
+
+
+class CompletedTextToImageResponse(TextToImageResponse):
+    """Narrowed response from ``run()`` once polling observes completion."""
+
+    images = required([lambda: Image])
+
+
+CompletedEditImageResponse = CompletedTextToImageResponse

runapi_seedream-0.1.0/tests/test_client.py

@@ -0,0 +1,145 @@
+import pytest
+
+from runapi.core import config
+from runapi.core.errors import AuthenticationError, ValidationError
+from runapi.seedream import SeedreamClient
+from runapi.seedream.resources.edit_image import EditImage
+from runapi.seedream.resources.text_to_image import TextToImage
+from runapi.seedream.types import CompletedTextToImageResponse, TextToImageResponse
+
+
+class FakeHttp:
+    def __init__(self, *responses):
+        self._responses = list(responses)
+        self.calls = []
+
+    def request(self, method, path, body=None, options=None):
+        self.calls.append((method, path, body))
+        if self._responses:
+            return self._responses.pop(0)
+        return {"id": "task_1", "status": "pending"}
+
+
+@pytest.fixture(autouse=True)
+def reset_config(monkeypatch):
+    monkeypatch.delenv("RUNAPI_API_KEY", raising=False)
+    monkeypatch.setattr(config, "api_key", None)
+    yield
+
+
+# --- auth -----------------------------------------------------------------
+
+
+def test_accepts_api_key_parameter():
+    assert isinstance(SeedreamClient(api_key="k", http_client=FakeHttp()), SeedreamClient)
+
+
+def test_falls_back_to_global(monkeypatch):
+    monkeypatch.setattr(config, "api_key", "global-key")
+    assert isinstance(SeedreamClient(http_client=FakeHttp()), SeedreamClient)
+
+
+def test_falls_back_to_env(monkeypatch):
+    monkeypatch.setenv("RUNAPI_API_KEY", "env-key")
+    assert isinstance(SeedreamClient(http_client=FakeHttp()), SeedreamClient)
+
+
+def test_raises_without_api_key():
+    with pytest.raises(AuthenticationError, match="API key is required"):
+        SeedreamClient()
+
+
+# --- injection / accessors ------------------------------------------------
+
+
+def test_uses_injected_http_client():
+    fake = FakeHttp()
+    client = SeedreamClient(api_key="k", http_client=fake)
+    assert client.text_to_image._http is fake
+    assert client.edit_image._http is fake
+
+
+def test_exposes_resource_accessors():
+    client = SeedreamClient(api_key="k", http_client=FakeHttp())
+    assert isinstance(client.text_to_image, TextToImage)
+    assert isinstance(client.edit_image, EditImage)
+
+
+# --- request shapes -------------------------------------------------------
+
+
+def test_create_posts_compacted_body():
+    fake = FakeHttp({"id": "t1", "status": "pending"})
+    client = SeedreamClient(api_key="k", http_client=fake)
+    result = client.text_to_image.create(
+        model="seedream-v4-text-to-image", prompt="hello world", aspect_ratio="1:1", seed=None
+    )
+    assert fake.calls == [
+        ("post", "/api/v1/seedream/text_to_image", {"model": "seedream-v4-text-to-image", "prompt": "hello world", "aspect_ratio": "1:1"}),
+    ]
+    assert isinstance(result, TextToImageResponse)
+
+
+def test_get_fetches_by_id():
+    fake = FakeHttp({"id": "t1", "status": "processing"})
+    client = SeedreamClient(api_key="k", http_client=fake)
+    client.text_to_image.get("t1")
+    assert fake.calls == [("get", "/api/v1/seedream/text_to_image/t1", None)]
+
+
+def test_run_narrows_completed_type():
+    fake = FakeHttp(
+        {"id": "t1", "status": "pending"},
+        {"id": "t1", "status": "completed", "images": [{"url": "https://x/y.png"}]},
+    )
+    client = SeedreamClient(api_key="k", http_client=fake)
+    result = client.text_to_image.run(model="seedream-v4-text-to-image", prompt="a serene lake")
+    assert isinstance(result, CompletedTextToImageResponse)
+    assert result.images[0].url == "https://x/y.png"
+
+
+# --- validation -----------------------------------------------------------
+
+
+def test_rejects_unknown_model():
+    client = SeedreamClient(api_key="k", http_client=FakeHttp())
+    with pytest.raises(ValidationError, match="model must be one of:"):
+        client.text_to_image.create(model="nope", prompt="hi there")
+
+
+def test_requires_prompt():
+    client = SeedreamClient(api_key="k", http_client=FakeHttp())
+    with pytest.raises(ValidationError, match="prompt is required"):
+        client.text_to_image.create(model="seedream-v4-text-to-image")
+
+
+def test_v4_output_count_range():
+    client = SeedreamClient(api_key="k", http_client=FakeHttp())
+    with pytest.raises(ValidationError, match="output_count must be one of: 1, 2, 3, 4, 5, 6"):
+        client.text_to_image.create(model="seedream-v4-text-to-image", prompt="hi there", output_count=9)
+
+
+def test_non_v4_requires_aspect_ratio_and_quality():
+    client = SeedreamClient(api_key="k", http_client=FakeHttp())
+    with pytest.raises(ValidationError, match="aspect_ratio is required"):
+        client.text_to_image.create(model="seedream-4.5-text-to-image", prompt="hi there")
+
+
+def test_text_to_image_rejects_source_image_urls():
+    client = SeedreamClient(api_key="k", http_client=FakeHttp())
+    with pytest.raises(ValidationError, match="source_image_urls is not supported"):
+        client.text_to_image.create(
+            model="seedream-v4-text-to-image", prompt="hi there", source_image_urls=["https://x/a.png"]
+        )
+
+
+def test_edit_requires_source_image_urls():
+    client = SeedreamClient(api_key="k", http_client=FakeHttp())
+    with pytest.raises(ValidationError, match="source_image_urls is required"):
+        client.edit_image.create(model="seedream-v4-edit", prompt="make it pop")
+
+
+def test_lite_prompt_min_length():
+    client = SeedreamClient(api_key="k", http_client=FakeHttp())
+    with pytest.raises(ValidationError, match="between 3 and 3000"):
+        client.text_to_image.create(model="seedream-5-lite-text-to-image", prompt="hi", aspect_ratio="1:1", output_quality="high")