@weirdfingers/baseboards 0.6.1 → 0.7.0

package/templates/api/src/boards/generators/implementations/fal/video/kling_video_ai_avatar_v2_standard.py

@@ -0,0 +1,159 @@
+"""
+fal.ai Kling Video AI Avatar v2 Standard generator.
+
+Generates avatar videos by synthesizing realistic humans, animals, cartoons, or
+stylized characters. Takes an image and audio as inputs to create synchronized
+video output.
+
+Based on Fal AI's fal-ai/kling-video/ai-avatar/v2/standard model.
+See: https://fal.ai/models/fal-ai/kling-video/ai-avatar/v2/standard
+"""
+
+import os
+
+from pydantic import BaseModel, Field
+
+from ....artifacts import AudioArtifact, ImageArtifact
+from ....base import BaseGenerator, GeneratorExecutionContext, GeneratorResult
+
+
+class KlingVideoAiAvatarV2StandardInput(BaseModel):
+    """Input schema for Kling Video AI Avatar v2 Standard.
+
+    Artifact fields are automatically detected via type introspection
+    and resolved from generation IDs to artifact objects.
+    """
+
+    image: ImageArtifact = Field(description="The image to use as your avatar")
+    audio: AudioArtifact = Field(description="The audio file for lip-sync animation")
+    prompt: str | None = Field(
+        default=".", description="The prompt to use for the video generation"
+    )
+
+
+class FalKlingVideoAiAvatarV2StandardGenerator(BaseGenerator):
+    """Generator for AI-powered avatar video synthesis."""
+
+    name = "fal-kling-video-ai-avatar-v2-standard"
+    description = "Fal: Kling Video AI Avatar v2 Standard - Avatar video from image and audio"
+    artifact_type = "video"
+
+    def get_input_schema(self) -> type[KlingVideoAiAvatarV2StandardInput]:
+        """Return the input schema for this generator."""
+        return KlingVideoAiAvatarV2StandardInput
+
+    async def generate(
+        self, inputs: KlingVideoAiAvatarV2StandardInput, context: GeneratorExecutionContext
+    ) -> GeneratorResult:
+        """Generate avatar video using fal.ai kling-video/ai-avatar/v2/standard."""
+        # Check for API key
+        if not os.getenv("FAL_KEY"):
+            raise ValueError("API configuration invalid. Missing FAL_KEY environment variable")
+
+        # Import fal_client
+        try:
+            import fal_client
+        except ImportError as e:
+            raise ImportError(
+                "fal.ai SDK is required for FalKlingVideoAiAvatarV2StandardGenerator. "
+                "Install with: pip install weirdfingers-boards[generators-fal]"
+            ) from e
+
+        # Upload image and audio artifacts to Fal's public storage
+        # Fal API requires publicly accessible URLs
+        from ..utils import upload_artifacts_to_fal
+
+        # Upload image and audio separately
+        image_urls = await upload_artifacts_to_fal([inputs.image], context)
+        audio_urls = await upload_artifacts_to_fal([inputs.audio], context)
+
+        # Prepare arguments for fal.ai API
+        arguments = {
+            "image_url": image_urls[0],
+            "audio_url": audio_urls[0],
+        }
+
+        # Add prompt only if provided and not the default empty value
+        if inputs.prompt and inputs.prompt != ".":
+            arguments["prompt"] = inputs.prompt
+
+        # Submit async job
+        handler = await fal_client.submit_async(
+            "fal-ai/kling-video/ai-avatar/v2/standard",
+            arguments=arguments,
+        )
+
+        # Store external job ID
+        await context.set_external_job_id(handler.request_id)
+
+        # Stream progress updates
+        from .....progress.models import ProgressUpdate
+
+        event_count = 0
+        async for event in handler.iter_events(with_logs=True):
+            event_count += 1
+            # Sample every 3rd event to avoid spam
+            if event_count % 3 == 0:
+                # Extract logs if available
+                logs = getattr(event, "logs", None)
+                if logs:
+                    # Join log entries into a single message
+                    if isinstance(logs, list):
+                        message = " | ".join(str(log) for log in logs if log)
+                    else:
+                        message = str(logs)
+
+                    if message:
+                        await context.publish_progress(
+                            ProgressUpdate(
+                                job_id=handler.request_id,
+                                status="processing",
+                                progress=50.0,  # Approximate mid-point progress
+                                phase="processing",
+                                message=message,
+                            )
+                        )
+
+        # Get final result
+        result = await handler.get()
+
+        # Extract video from result
+        # fal.ai returns: {"video": {"url": ..., "content_type": ...}, "duration": ...}
+        video_data = result.get("video")
+
+        if not video_data:
+            raise ValueError("No video returned from fal.ai API")
+
+        video_url = video_data.get("url")
+        if not video_url:
+            raise ValueError("Video missing URL in fal.ai response")
+
+        # Extract format from content_type (e.g., "video/mp4" -> "mp4")
+        content_type = video_data.get("content_type", "video/mp4")
+        video_format = content_type.split("/")[-1] if "/" in content_type else "mp4"
+
+        # Extract duration from response (available in API response)
+        duration = result.get("duration")
+
+        # Store the video result
+        # Use input image dimensions as base dimensions (avatar video maintains aspect ratio)
+        artifact = await context.store_video_result(
+            storage_url=video_url,
+            format=video_format,
+            width=inputs.image.width,
+            height=inputs.image.height,
+            duration=duration,
+            fps=None,  # Not provided by API
+            output_index=0,
+        )
+
+        return GeneratorResult(outputs=[artifact])
+
+    async def estimate_cost(self, inputs: KlingVideoAiAvatarV2StandardInput) -> float:
+        """Estimate cost for Kling Video AI Avatar v2 Standard generation in USD.
+
+        Pricing not specified in documentation, using estimate based on
+        typical Kling video generation costs.
+        """
+        # Estimate per generation based on typical Kling pricing
+        return 0.10

package/templates/api/src/boards/generators/implementations/fal/video/veed_fabric_1_0.py

@@ -0,0 +1,180 @@
+"""
+VEED Fabric 1.0 image-to-video generator.
+
+Generate talking videos from any image using VEED Fabric 1.0.
+This generator turns a static image into a talking video with synchronized
+lip movements based on provided audio.
+
+Based on Fal AI's veed/fabric-1.0 model.
+See: https://fal.ai/models/veed/fabric-1.0
+"""
+
+import os
+from typing import Literal
+
+from pydantic import BaseModel, Field
+
+from ....artifacts import AudioArtifact, ImageArtifact
+from ....base import BaseGenerator, GeneratorExecutionContext, GeneratorResult
+from ..utils import upload_artifacts_to_fal
+
+
+class VeedFabric10Input(BaseModel):
+    """Input schema for VEED Fabric 1.0.
+
+    Artifact fields are automatically detected via type introspection
+    and resolved from generation IDs to artifact objects.
+    """
+
+    image_url: ImageArtifact = Field(description="Image to turn into a talking video")
+    audio_url: AudioArtifact = Field(description="Audio to synchronize with the image")
+    resolution: Literal["720p", "480p"] = Field(
+        default="720p", description="Output video resolution"
+    )
+
+
+class FalVeedFabric10Generator(BaseGenerator):
+    """Generator for turning images into talking videos using VEED Fabric 1.0."""
+
+    name = "veed-fabric-1.0"
+    description = "VEED: Fabric 1.0 - Turn any image into a talking video"
+    artifact_type = "video"
+
+    def get_input_schema(self) -> type[VeedFabric10Input]:
+        """Return the input schema for this generator."""
+        return VeedFabric10Input
+
+    async def generate(
+        self, inputs: VeedFabric10Input, context: GeneratorExecutionContext
+    ) -> GeneratorResult:
+        """Generate talking video using VEED Fabric 1.0."""
+        # Check for API key
+        if not os.getenv("FAL_KEY"):
+            raise ValueError("API configuration invalid. Missing FAL_KEY environment variable")
+
+        # Import fal_client
+        try:
+            import fal_client
+        except ImportError as e:
+            raise ImportError(
+                "fal.ai SDK is required for FalVeedFabric10Generator. "
+                "Install with: pip install weirdfingers-boards[generators-fal]"
+            ) from e
+
+        # Upload image and audio artifacts to Fal's public storage
+        # Fal API requires publicly accessible URLs
+        image_urls = await upload_artifacts_to_fal([inputs.image_url], context)
+        audio_urls = await upload_artifacts_to_fal([inputs.audio_url], context)
+
+        # Prepare arguments for fal.ai API
+        arguments = {
+            "image_url": image_urls[0],
+            "audio_url": audio_urls[0],
+            "resolution": inputs.resolution,
+        }
+
+        # Submit async job
+        handler = await fal_client.submit_async(
+            "veed/fabric-1.0",
+            arguments=arguments,
+        )
+
+        # Store external job ID
+        await context.set_external_job_id(handler.request_id)
+
+        # Stream progress updates
+        from .....progress.models import ProgressUpdate
+
+        event_count = 0
+        async for event in handler.iter_events(with_logs=True):
+            event_count += 1
+            # Sample every 3rd event to avoid spamming progress updates
+            if event_count % 3 == 0:
+                logs = getattr(event, "logs", None)
+                if logs:
+                    if isinstance(logs, list):
+                        message = " | ".join(str(log) for log in logs if log)
+                    else:
+                        message = str(logs)
+
+                    if message:
+                        await context.publish_progress(
+                            ProgressUpdate(
+                                job_id=handler.request_id,
+                                status="processing",
+                                progress=50.0,
+                                phase="processing",
+                                message=message,
+                            )
+                        )
+
+        # Get final result
+        result = await handler.get()
+
+        # Extract video from result
+        # Fabric 1.0 API returns: {"video": {"url": "...", "content_type": "video/mp4", ...}}
+        video_data = result.get("video")
+
+        if not video_data:
+            raise ValueError(
+                "No video returned from VEED Fabric 1.0 API. "
+                f"Response structure: {list(result.keys())}"
+            )
+
+        video_url = video_data.get("url")
+        if not video_url:
+            raise ValueError(
+                f"Video missing URL in VEED response. Video data keys: {list(video_data.keys())}"
+            )
+
+        # Determine video format with fallback strategy:
+        # 1. Try to extract from URL extension
+        # 2. Parse content_type only if it's a video/* MIME type
+        # 3. Default to mp4
+        video_format = "mp4"  # Default
+
+        if video_url:
+            url_parts = video_url.split(".")
+            if len(url_parts) > 1:
+                ext = url_parts[-1].split("?")[0].lower()
+                if ext in ["mp4", "webm", "mov", "avi"]:
+                    video_format = ext
+
+        if video_format == "mp4":
+            content_type = video_data.get("content_type", "")
+            if content_type.startswith("video/"):
+                video_format = content_type.split("/")[-1]
+
+        # Determine output dimensions based on resolution
+        # 720p = 1280x720, 480p = 854x480 (16:9 aspect ratio)
+        if inputs.resolution == "720p":
+            output_width = 1280
+            output_height = 720
+        else:
+            output_width = 854
+            output_height = 480
+
+        # Store the video result
+        artifact = await context.store_video_result(
+            storage_url=video_url,
+            format=video_format,
+            width=output_width,
+            height=output_height,
+            duration=inputs.audio_url.duration,
+            fps=30.0,  # Standard frame rate for talking videos
+            output_index=0,
+        )
+
+        return GeneratorResult(outputs=[artifact])
+
+    async def estimate_cost(self, inputs: VeedFabric10Input) -> float:
+        """Estimate cost for VEED Fabric 1.0 generation in USD.
+
+        Pricing not specified in documentation, using estimate based on
+        typical video lipsync processing costs.
+        """
+        # Fixed cost estimate of $0.08 per generation
+        # Based on typical AI video processing costs
+        # This is a conservative estimate and should be updated when official
+        # pricing information becomes available
+        return 0.08

package/templates/api/src/boards/generators/implementations/fal/video/veo31.py

@@ -0,0 +1,190 @@
+"""
+Google Veo 3.1 text-to-video generator.
+
+The most advanced AI video generation model in the world by Google, featuring
+audio generation capabilities and enhanced quality over Veo 3.
+
+Based on Fal AI's fal-ai/veo3.1 model.
+See: https://fal.ai/models/fal-ai/veo3.1
+"""
+
+import os
+from typing import Literal
+
+from pydantic import BaseModel, Field
+
+from ....base import BaseGenerator, GeneratorExecutionContext, GeneratorResult
+
+
+class Veo31Input(BaseModel):
+    """Input schema for Google Veo 3.1 text-to-video generation."""
+
+    prompt: str = Field(description="The text prompt describing the video you want to generate")
+    aspect_ratio: Literal["9:16", "16:9"] = Field(
+        default="16:9",
+        description="Aspect ratio of the generated video",
+    )
+    duration: Literal["4s", "6s", "8s"] = Field(
+        default="8s",
+        description="Duration of the generated video",
+    )
+    resolution: Literal["720p", "1080p"] = Field(
+        default="720p",
+        description="Resolution of the generated video",
+    )
+    generate_audio: bool = Field(
+        default=True,
+        description="Whether to generate audio for the video. If false, 50% less credits used",
+    )
+    enhance_prompt: bool = Field(
+        default=True,
+        description="Whether to enhance video generation quality",
+    )
+    auto_fix: bool = Field(
+        default=True,
+        description="Automatically attempt to rewrite prompts that fail content policy validation",
+    )
+    seed: int | None = Field(
+        default=None,
+        description="Seed value for reproducible generation",
+    )
+    negative_prompt: str | None = Field(
+        default=None,
+        description="Guidance text to exclude elements from generation",
+    )
+
+
+class FalVeo31Generator(BaseGenerator):
+    """Generator for text-to-video using Google Veo 3.1."""
+
+    name = "fal-veo31"
+    description = "Fal: Veo 3.1 - Google's most advanced AI video generation model with audio"
+    artifact_type = "video"
+
+    def get_input_schema(self) -> type[Veo31Input]:
+        """Return the input schema for this generator."""
+        return Veo31Input
+
+    async def generate(
+        self, inputs: Veo31Input, context: GeneratorExecutionContext
+    ) -> GeneratorResult:
+        """Generate video using fal.ai veo3.1."""
+        # Check for API key
+        if not os.getenv("FAL_KEY"):
+            raise ValueError("API configuration invalid. Missing FAL_KEY environment variable")
+
+        # Import fal_client
+        try:
+            import fal_client
+        except ImportError as e:
+            raise ImportError(
+                "fal.ai SDK is required for FalVeo31Generator. "
+                "Install with: pip install weirdfingers-boards[generators-fal]"
+            ) from e
+
+        # Prepare arguments for fal.ai API
+        arguments = {
+            "prompt": inputs.prompt,
+            "aspect_ratio": inputs.aspect_ratio,
+            "duration": inputs.duration,
+            "resolution": inputs.resolution,
+            "generate_audio": inputs.generate_audio,
+            "enhance_prompt": inputs.enhance_prompt,
+            "auto_fix": inputs.auto_fix,
+        }
+
+        # Add optional parameters if provided
+        if inputs.seed is not None:
+            arguments["seed"] = inputs.seed
+        if inputs.negative_prompt is not None:
+            arguments["negative_prompt"] = inputs.negative_prompt
+
+        # Submit async job
+        handler = await fal_client.submit_async(
+            "fal-ai/veo3.1",
+            arguments=arguments,
+        )
+
+        # Store external job ID
+        await context.set_external_job_id(handler.request_id)
+
+        # Stream progress updates
+        from .....progress.models import ProgressUpdate
+
+        event_count = 0
+        async for event in handler.iter_events(with_logs=True):
+            event_count += 1
+            # Sample every 3rd event to avoid spam
+            if event_count % 3 == 0:
+                # Extract logs if available
+                logs = getattr(event, "logs", None)
+                if logs:
+                    # Join log entries into a single message
+                    if isinstance(logs, list):
+                        message = " | ".join(str(log) for log in logs if log)
+                    else:
+                        message = str(logs)
+
+                    if message:
+                        await context.publish_progress(
+                            ProgressUpdate(
+                                job_id=handler.request_id,
+                                status="processing",
+                                progress=50.0,
+                                phase="processing",
+                                message=message,
+                            )
+                        )
+
+        # Get final result
+        result = await handler.get()
+
+        # Extract video from result
+        # Expected structure: {"video": {"url": "...", "content_type": "...", ...}}
+        video_data = result.get("video")
+        if not video_data:
+            raise ValueError("No video returned from fal.ai API")
+
+        video_url = video_data.get("url")
+        if not video_url:
+            raise ValueError("Video missing URL in fal.ai response")
+
+        # Determine video dimensions based on resolution and aspect ratio
+        if inputs.resolution == "720p":
+            if inputs.aspect_ratio == "16:9":
+                width, height = 1280, 720
+            else:  # 9:16
+                width, height = 720, 1280
+        else:  # 1080p
+            if inputs.aspect_ratio == "16:9":
+                width, height = 1920, 1080
+            else:  # 9:16
+                width, height = 1080, 1920
+
+        # Parse duration from "8s" format
+        duration_seconds = int(inputs.duration.rstrip("s"))
+
+        # Store video result
+        artifact = await context.store_video_result(
+            storage_url=video_url,
+            format="mp4",
+            width=width,
+            height=height,
+            duration=duration_seconds,
+            output_index=0,
+        )
+
+        return GeneratorResult(outputs=[artifact])
+
+    async def estimate_cost(self, inputs: Veo31Input) -> float:
+        """Estimate cost for this generation in USD.
+
+        Note: Pricing information not available in Fal documentation.
+        Using placeholder value that should be updated with actual pricing.
+        """
+        # TODO: Update with actual pricing from Fal when available
+        # Base cost, with 50% reduction if audio is disabled
+        base_cost = 0.15  # Placeholder estimate
+        if not inputs.generate_audio:
+            return base_cost * 0.5
+        return base_cost