Model Catalog
Full transparency. The AI models that power your screenplay editor, scene builder, and video workflows—writer-directed and production-aware.
LLM Models
12 models in the editor (Story Advisor, Character, Location agents + modals)
| Name | Provider | Description |
|---|---|---|
| Claude Haiku 4.5 | Anthropic | Fastest and most cost-efficient Claude. Near-frontier performance for real-time and high-volume use. |
| Claude Opus 4.6 | Anthropic | Anthropic's most capable model. Adaptive thinking, 1M context, leads on complex reasoning benchmarks. |
| Claude Sonnet 4.6 | Anthropic | Anthropic's most capable Sonnet. Strong at creative writing, coding, and agent planning. 1M token context in beta. |
| Gemini 2.5 Flash | Best price/performance. First Flash with thinking. Fast, well-rounded, multimodal. | |
| Gemini 2.5 Pro | Advanced reasoning with thinking. Excels at complex code, math, STEM, and long-context analysis. | |
| Gemini 3 Pro | Google's most intelligent model. Natively multimodal, 1M context, state-of-the-art reasoning. | |
| GPT-4o | OpenAI | Multimodal flagship (text, audio, image, video). Fast, strong vision and instruction following. |
| GPT-5.4 | OpenAI | Frontier reasoning model for complex professional work with stronger precision on hard writing and planning tasks. |
| Grok 4.1 Fast | xAI | Low latency with reasoning for fast analysis loops. |
| Grok 4.1 Fast Lite | xAI | Speed-prioritized fast mode for lightweight drafting and rewrites. |
| Grok 4.20 | xAI | Advanced reasoning model with 2M context. Best for deep analysis and planning. |
| O3 | OpenAI | Reasoning-focused. Step-by-step logic for complex coding, math, and science. Agentic tool use. |
Image Models
6 models exposed across Scene Builder and image generation tools
| Name | Provider | Description |
|---|---|---|
| FLUX.2 [max] | Black Forest Labs | Top-tier FLUX.2 model. 2K/4K tiers, character consistency, up to 10 refs. Cinematic visuals and production typography. |
| FLUX.2 [pro] | Black Forest Labs | Production-grade FLUX.2 model. 2K/4K tiers, up to 8-10 refs. Photorealistic detail, hex color control, spatial reasoning. |
| Grok Imagine (Standard) | xAI | Fast text-to-image and editing model. Supports up to 5 input images for edits, 2K output, and higher throughput (300 RPM). |
| Grok Imagine Pro | xAI | Higher-tier Grok Imagine for quality-focused shots. Supports text+image editing, 2K output, with lower-volume premium throughput (30 RPM). |
| Nano Banana Pro | Google DeepMind image model. Up to 2K native, 4K upscale. Strong text rendering, up to 14 reference images. | |
| Nano Banana Pro2 | Gemini 3.1 Flash Image model. Standard and premium resolution tiers, strong editing/text rendering, up to 14 reference images in Wryda workflows. |
Video Models
9 models in the Video Gen dropdown (playground)
| Name | Provider | Description |
|---|---|---|
| Google Veo 3.1 (Quality) | Highest quality. Native audio, lip-sync, multi-person dialogue. Reference images, scene extension. | |
| Google Veo 3.1 Fast | Faster generation, slightly lower quality. Native audio, reference images, scene extension. | |
| LTX 2.3 Fast | Lightricks | Optimized for fast, lower-cost text/image-to-video iteration; supports longer 1080p durations up to 20s. |
| LTX 2.3 Pro | Lightricks | Higher-fidelity LTX output with improved motion stability and detail; required for audio-to-video, retake, and extend workflows. |
| Luma Ray 2 | Luma | Photorealistic video from text. 5 or 9 second clips, keyframes, camera controls, and 4K output support. |
| Luma Ray 2 Flash | Luma | Faster, ~1/3 cost of Ray 2. Natural motion, keyframes, up to 9 seconds at 4K. |
| Runway Gen-4 Turbo | Runway | Fast image-to-video. ~5x faster than Gen-4. 10-second clips in ~30 seconds. |
| Runway Gen-4.5 | Runway | Top-rated text-to-video. Cinematic, photorealistic. Physics-accurate motion, fine detail preservation. |
| xAI Grok Video | xAI | Text and image to video. Editing, restyling, motion control. Supports 5, 10, and 15 second outputs. |