AI Models & Specifications
Dal Nulla leverages cutting-edge AI models to bring your creative visions to life. Here’s a breakdown of the models we use and their key capabilities.
Quick Summary
Text
We use the latest generation of Gemini models to power our text generation features, ensuring high-quality reasoning and creative writing capabilities.
Gemini 3 Pro Preview
Description: A state-of-the-art reasoning model designed for complex tasks. It excels at understanding intricate instructions, maintaining long-context coherence, and generating nuanced creative text.
- Reasoning: Advanced logic and reasoning capabilities for sophisticated prompt engineering and scriptwriting.
- Context: Large context window to maintain consistency across long narratives or detailed project briefs.
- Use Cases: Ideal for generating complex scripts, detailed character backstories, and intricate scene descriptions.
Gemini 3 Flash Preview
Description: A highly optimized model built for speed and efficiency without sacrificing significant quality. It is perfect for rapid ideation and quick text generation tasks.
- Speed: Extremely fast response times for real-time interaction and iteration.
- Efficiency: Optimized for high-volume tasks and quick brainstorming sessions.
- Use Cases: Great for generating quick scene ideas, dialogue snippets, and alternative prompt variations.
Image
We use Gemini 2.5 Flash for all our image generation and editing needs, providing high-resolution outputs and versatile editing capabilities.
Gemini 2.5 Flash (for Image Generation & Editing)
Alias: nanobanana
Gemini 2.5 Flash is our primary model for all static image generation and AI-powered image editing tasks. It’s optimized for high-fidelity visual output and speed.
- Resolution: Generates and edits images up to 4K resolution (3840x2160 pixels).
- Speed: Designed for rapid iteration and quick results.
- Editing Capabilities: Supports various AI editing functions, including inpainting, outpainting, style transfer, and content-aware manipulation.
- Versatility: Excellent for diverse artistic styles, photographic realism, and abstract compositions.
Video
Veo 3.1 is our powerful suite of models dedicated to video generation. It offers different modes to balance speed, quality, and duration.
Veo 3.1 Flash
Description: Optimized for fast video generation, ideal for quickly prototyping ideas or creating shorter, dynamic clips.
- Max Duration: Generates videos up to 150 seconds (2 minutes and 30 seconds).
- Resolution: Supports various resolutions, including up to 4K (3840x2160 pixels) for videos up to 8 seconds. Videos longer than 8 seconds are automatically rendered at 720p to ensure optimal performance and stability.
- Audio Integration: Automatically synchronizes generated audio with video content.
- Speed: High-speed processing for rapid video output.
Veo 3.1 Quality
Description: Prioritizes high-fidelity video output, offering enhanced visual detail and coherence, suitable for more polished productions. This model is generally slower but produces superior visual quality.
- Max Duration: Generates videos up to 150 seconds (2 minutes and 30 seconds).
- Resolution: Supports various resolutions, including up to 4K (3840x2160 pixels) for videos up to 8 seconds. Videos longer than 8 seconds are automatically rendered at 720p to ensure optimal performance and stability.
- Audio Integration: Automatically synchronizes generated audio with video content.
- Quality: Produces high-fidelity visuals with improved temporal consistency and detail.