Runway Launches Gen-4.5: Now the #1 Ranked AI Video Generator
Key Takeaways
- âś“ Runway Gen-4.5 now ranks #1 on the Artificial Analysis Text-to-Video benchmark with 1,247 Elo points
- âś“ Significant advances in motion quality, visual fidelity, and prompt adherence
- âś“ Maintains the speed and efficiency of Gen-4 while delivering breakthrough quality
- âś“ Available at same pricing as previous models across all subscription plans
- âś“ Built entirely on NVIDIA GPUs with inference on Hopper and Blackwell series
- âś“ All existing control modes (Image-to-Video, Keyframes, Video-to-Video) coming to Gen-4.5
What Happened
Runway announced the release of Gen-4.5 on December 1, 2025—their new frontier video generation model that now holds the top position in AI video generation benchmarks.
With 1,247 Elo points, Gen-4.5 currently leads the Artificial Analysis Text-to-Video benchmark, surpassing all other models including OpenAI’s Sora, Google’s Veo 2, and other competitors.
“Gen-4.5 pushes the frontier of video generation even further. It represents significant advances in both pre-training data efficiency and post-training techniques.” — Runway Research Team
Key Capabilities
Precise Prompt Adherence
Gen-4.5 achieves unprecedented physical accuracy and visual precision. The model excels at:
- Complex Scenes: Intricate, multi-element scenes rendered with precision
- Physical Accuracy: Realistic physics with believable collisions and natural movement
- Detailed Compositions: Precise placement and fluid motion for objects and characters
- Expressive Characters: Nuanced emotions, natural gestures, and lifelike facial detail
| Capability | Description |
|---|---|
| Motion Quality | Natural physics, fluid movements, realistic human actions |
| Prompt Adherence | Complex multi-step instructions executed with high fidelity |
| Visual Fidelity | Objects move with believable weight, momentum, and force |
| Temporal Consistency | Maintains coherence across frames and scenes |
Stylistic Control
Gen-4.5 supports a wide range of aesthetics while maintaining a coherent visual language:
- Photorealistic: Visuals indistinguishable from real-world footage
- Non-photorealistic: Stylized, expressive motion with artistic freedom
- Cinematic: Emotionally powerful visuals with striking depth
- Slice of Life: Everyday scenes with authentic, true-to-life detail
Technical Foundation
Gen-4.5 was developed entirely on NVIDIA GPUs across initial R&D, pre-training, post-training, and inference. Runway collaborated extensively with NVIDIA to optimize video diffusion models.
“We’re proud that Runway built their groundbreaking video and world model on NVIDIA GPUs, and are thrilled to see Runway revolutionize the video generation industry.” — Jensen Huang, President and CEO of NVIDIA
Inference runs on NVIDIA Hopper and Blackwell series GPUs, delivering optimized performance without compromising quality.
Explore AI Video Generation
Compare the best AI video generators for your creative projects
See Our Comparisons →Known Limitations
Despite the significant advances, Runway acknowledges several limitations common to video generation models:
- Causal reasoning: Effects sometimes precede causes (e.g., a door opening before the handle is pressed)
- Object permanence: Objects may disappear or appear unexpectedly across frames
- Success bias: Actions disproportionately succeed (e.g., a poorly aimed kick still scoring a goal)
These limitations are particularly important for world models that need to accurately represent action outcomes.
Why This Matters for Creators
For Content Creators
Gen-4.5 delivers Hollywood-grade motion and visual fidelity for short product narratives, social media content, and stylized brand worlds—especially when multiple variants are needed fast.
For Filmmakers
The model excels at storyboards, previsualization, and animatics. Creators can explore scene blocking, camera direction, mood, and transitions before committing to production.
For VFX Artists
Rapidly test atmosphere, lighting changes, background swaps, or shot concepts, then hand off the best version to a traditional pipeline.
Pricing and Availability
Gen-4.5 maintains the speed and efficiency of Gen-4, delivering breakthrough quality without compromising performance. The model is available at comparable pricing across all subscription plans, making world-leading video generation accessible to creators and organizations at every scale.
All existing control modes—Image-to-Video, Keyframes, Video-to-Video, and more—will be brought to Gen-4.5.
The Competitive Landscape
Gen-4.5 enters a heated race in AI video generation:
- OpenAI Sora: High-fidelity video with Disney character licensing
- Google Veo 2: Advanced physics simulations and 3D consistency
- Pika Labs: Stylized video generation with rapid iteration
- Kling: Motion capabilities and extended video lengths
Runway’s differentiation lies in its benchmark-leading quality combined with accessible pricing and a comprehensive creative toolset.
What we’re watching: How competitors respond to Gen-4.5’s benchmark leadership, and whether Runway’s quality advantages translate to broader adoption among professional creators.
Sources
- Runway Official Announcement
- CNBC: Runway Gen 4.5 video model beats Google, OpenAI
- Artificial Analysis Text-to-Video Benchmark