gif_generator.py

from google import genai
from google.genai import types
from PIL import Image
from io import BytesIO
import os
import subprocess
import tempfile
from global_config import global_config
from loguru import logger as log

# TODO: Use your own key
GEMINI_API_KEY = ""

# Setup logging
setup_logging()

client = genai.Client(api_key=GEMINI_API_KEY)

contents = "Create an animation by generating multiple frames, showing a seed growing into a plant and then blooming into a flower, in a pixel art style"

log.info(f"Sending request to Gemini with prompt: {contents}")
response = client.models.generate_content(
    model="models/gemini-2.0-flash-exp",
    contents=contents,
    config=types.GenerateContentConfig(response_modalities=['Text', 'Image'])
)

# Create a temporary directory to store the frames
with tempfile.TemporaryDirectory() as temp_dir:
    log.info(f"Created temporary directory at {temp_dir}")
    frame_paths = []
    frame_count = 0
    
    # Process and save each part
    log.info(f"Number of candidates: {len(response.candidates)}")
    if response.candidates:
        log.info(f"Number of parts in first candidate: {len(response.candidates[0].content.parts)}")
        
        for part_index, part in enumerate(response.candidates[0].content.parts):
            log.info(f"Processing part {part_index+1}")
            if part.text is not None:
                log.info(f"Text content: {part.text[:100]}...")
                print(part.text)
            elif part.inline_data is not None:
                # Save the image to a temporary file
                frame_path = os.path.join(temp_dir, f"frame_{frame_count:03d}.png")
                image = Image.open(BytesIO(part.inline_data.data))
                image.save(frame_path)
                frame_paths.append(frame_path)
                frame_count += 1
                log.info(f"Saved frame {frame_count} to {frame_path}")
            else:
                log.warning(f"Part {part_index+1} has neither text nor inline_data")
    else:
        log.error("No candidates returned in the response")
    
    # If we have frames, create a GIF using ffmpeg
    if frame_paths:
        log.info(f"Found {len(frame_paths)} frames to process")
        output_path = os.path.abspath("animation.gif")
        log.info(f"Will save animation to {output_path}")
        
        # List all files in the temp directory to verify
        log.info(f"Files in temp directory: {os.listdir(temp_dir)}")
        
        # Build ffmpeg command
        ffmpeg_cmd = [
            "ffmpeg",
            "-y",  # Overwrite output file if it exists
            "-framerate", "2",  # 2 frames per second
            "-pattern_type", "glob",
            "-i", f"{temp_dir}/frame_*.png",
            "-vf", "scale=512:-1:flags=lanczos",  # Resize while maintaining aspect ratio
            output_path
        ]
        
        try:
            cmd_str = ' '.join(ffmpeg_cmd)
            log.info(f"Running ffmpeg command: {cmd_str}")
            
            # Run ffmpeg and capture output
            result = subprocess.run(
                ffmpeg_cmd, 
                check=True,
                capture_output=True,
                text=True
            )
            
            log.info(f"ffmpeg stdout: {result.stdout}")
            log.info(f"ffmpeg stderr: {result.stderr}")
            
            if os.path.exists(output_path):
                log.info(f"Animation successfully saved to {output_path}")
                file_size = os.path.getsize(output_path)
                log.info(f"File size: {file_size} bytes")
                
                # Open the resulting GIF
                Image.open(output_path).show()
            else:
                log.error(f"Output file {output_path} was not created")
        except subprocess.CalledProcessError as e:
            log.error(f"Failed to create GIF: {e}")
            log.error(f"ffmpeg stdout: {e.stdout}")
            log.error(f"ffmpeg stderr: {e.stderr}")
        except Exception as e:
            log.error(f"Unexpected error: {str(e)}")
    else:
        log.warning("No frames were generated, cannot create animation")

log.info("Script completed")

It only generates text. Do I need to enable something in my API Key in the Google Console?

Ah sorry better refer to my newer one: https://x.com/Eito_Miyamura/status/1900190111565836455

This one little bit unstable