The latest AI news and updates of Runway — AI creative platform for video generation, Characters API, and General World Models. Covering Runway's latest product updates, launches, and research from the past 90 days.
Runway21h agoRunway integrates with ChatGPT via the Model Context Protocol, enabling image and video generation directly within conversations. The integration provides access to models including Seedance 2.5 and Kling without requiring an API key. Generations consume existing Runway credits and save automatically to the Runway library, with support also available for Claude, Cursor, and Replit.
Runway21h agoRunway now offers enterprise licensing for its state-of-the-art model weights. This program provides companies with model weights, training scripts, and white-glove support from researchers to fine-tune models on proprietary data. Licensed models run entirely within an enterprise's own infrastructure, allowing for full control over deployment, versioning, and commercialization of the resulting outputs.
RunwaySep 10Runway published a research overview detailing its approach to real-time video generation, prioritizing reduced time-to-first-frame and streaming outputs. The method converts base models into causal, autoregressive generators, using two-stage distillation to achieve low-latency performance. This research aims to enable interactive applications in gaming, robotics, and simulation by moving beyond static, multi-stage generation processes.
RunwaySep 9Runway launched free plugins for Adobe Premiere Pro and After Effects, integrating AI generation, editing, and upscaling directly into the timeline. The panel supports video and image prompting, re-rendering, and subject isolation, with results landing at the playhead. Generations consume existing Runway credits and include access to models like Gen-4.5, Veo 3.1, and Kling 3.0 Pro.
RunwaySep 4Runway launched a self-serve Team Plan for 2 to 9 members, featuring a pooled credit balance and shared projects. The plan includes frame-accurate commenting, asset lineage, 1TB of storage, and access to custom agent skills and Agent Connectors. It costs $55 per seat monthly when billed annually, providing 165,600 credits per year for the entire team.
RunwaySep 3Runway released GWM Worlds 2, a General World Model generating interactive, real-time simulations at 720p and 24 fps with 48000 Hz audio. The model supports defined environments, subjects, and physical rules, responding to text actions and continuous camera motion. It uses a WorldPrompt format to maintain persistent world states, enabling continuous, non-scripted sessions for entertainment and robotics simulation.
RunwaySep 2Runway launched Runway Dev MCP, a tool connecting its developer platform directly to coding agents. This integration allows agents to build, debug, and manage media generation workflows, including setting up Model Routers and looking up tasks, without leaving the editor. The platform is available now at dev.runway.com.
RunwaySep 1Runway now supports ACES (Academy Color Encoding System) exports within its Ruby model. The model generates scene-referred, half-float EXR sequences in ACEScg 1.3 and 2.0 formats. These outputs enable direct integration of Ruby’s SDR-to-HDR video conversions into professional color-managed post-production pipelines.
RunwayAug 31Runway has introduced Solaris, an Interface World Model that generates interactive software interfaces frame-by-frame in real-time without code. By eliminating intermediate representations like code, the model synthesizes interfaces directly from user interactions. Solaris outperformed frontier LLMs in structural similarity and information retention during interface reconstruction benchmarks. Interested parties can request early access through the Runway website.
RunwayAug 29Runway added Google's Omni 1.1 Flash video generation and editing model to its creative platform. The model supports text-to-video, image-to-video, and video-to-video generation. Access is available now directly within the Runway app.
RunwayAug 26Runway adds Meta's Muse Image model to its creative platform today. The model joins the platform's existing image and video generation tools, supporting image generation directly within the Runway interface. Access to the model is available now through the Runway app.
RunwayAug 25Runway has added Alibaba's WAN 3.0 model to its creative platform. The model generates video and audio content using multiple image, video, and audio reference inputs. WAN 3.0 is available now for use directly within the Runway interface.
RunwayAug 22Runway released Ruby, a new model that converts SDR video into 16-bit HDR. It outputs 16-bit EXR sequences or 10- and 12-bit ProRes and HEVC in BT.2020 color with PQ or HLG. The model supports any uploaded or generated video up to 30 seconds and is available now for Max and Enterprise plan subscribers.
RunwayAug 22Runway provides unlimited generation access to the MiniMax H3 video model for all Max plan subscribers. This offer removes usage caps and generation limits for the model until August 23.
RunwayAug 13Runway integrated xAI's Grok Imagine Image 2.0 model into its creative platform. The model joins the platform's existing image and video generation tools, providing additional options for generative media production. Access to the model is available now through the Runway interface.
RunwayAug 12Runway added connectors for Figma, Dropbox, and Notion to Agent on Runway. These integrations sync designs, files, and documents directly into the agent workspace, centralizing assets from external platforms. This feature is available to all users on paid plans.
RunwayAug 12Runway added the LTX-2.5 video generation model to its creative platform today. Developed by Lightricks, the model supports both text-to-video and image-to-video generation. It is now available for use directly within the Runway interface.
RunwayAug 5Runway is adding ByteDance’s Seedance 2.5 video model to its creative platform. The model supports up to 50 multimodal reference inputs and generates native video clips up to 30 seconds long. New Max plan subscribers receive 7 days of unlimited access to the model upon its upcoming launch.
RunwayAug 1Runway integrated xAI’s Grok Imagine Video 1.5 model into its creative platform. The model generates video from still images, allowing for motion and camera control directly within the Runway interface. This addition expands the platform's available generative media tools, providing access to xAI’s latest image-to-video capabilities for video production workflows.
RunwayJul 31Runway added the MiniMax H3 model to its creative platform. This open-weight multimodal model supports text, image, audio, and video inputs with native audio-visual output. It enables instruction-led edits, brand rendering, and video-to-video motion transfer. Users can now access H3 alongside other frontier models directly within the Runway interface.
RunwayJul 24Runway launched Workflows within Runway Agent, enabling the creation, execution, and editing of node-based workflows through natural language. The system triggers these pipelines via the /Workflow slash command. This update automates complex creative production by allowing direct, conversational control over node-based processes.
RunwayJul 23Runway launched Runway Media Router, a preference-optimized routing layer for generative media. Developers define priorities for cost, quality, or latency once, and the system automatically selects the appropriate video, image, or audio model for each request. The router is available now within the Runway Dev platform to streamline production media generation pipelines.
RunwayJul 17Runway Agent 2.0 ranked first across all three pillars—Narrative Coherence, Cinematic Language, and Production Quality—in Physion Labs’ independent Arc 1.0 benchmark. The evaluation tested six AI video agents against 30 cinematic prompts and 16 weighted metrics. Runway led every evaluation dimension, with its advantage particularly clear across subjective metrics where cinematic taste is prioritized.
RunwayJul 9Runway integrated ByteDance’s Seedream 5.0 Pro image generation model into its creative platform. The model generates detailed images from text prompts or reference inputs and supports native, legible text rendering in up to 14 languages. This integration provides tools for creating text-heavy visual content like posters and infographics directly within the Runway interface.
RunwayJul 8Runway launched Runway Dev, a unified media platform providing a single API for image, video, audio, and interactive character models. The platform includes pre-built Recipes for common creative outcomes, custom Workflows for pipeline automation, and exclusive real-time character avatars. It features enterprise-grade security, including SOC 2 Type II compliance and contractual data-retention controls.
RunwayJul 2Runway launched Agent Skills, a new capability for its AI agent that automates marketing tasks. A slash-command interface triggers the agent to generate ad campaigns, commercials, and localized assets. This system executes end-to-end creative workflows to scale marketing content production on command.
RunwayJul 1Runway now hosts Google’s Nano Banana 2 Lite, a speed-optimized model for image generation. The integration provides high-speed image creation directly within the Runway platform or via the Runway Agent. This model generates images at high speed without compromising quality, offering a new option for rapid visual content production.
RunwayJun 30Runway now supports Google's Gemini Omni Flash for video generation and editing. The integration processes text prompts, images, or existing videos to create content. Access is available directly through the Runway platform or by requesting the Runway Agent to use Omni for creative tasks.
RunwayJun 30Runway announced a strategic partnership with Japanese entertainment company MIXI to deploy its generative AI platform across the organization. The collaboration focuses on applying world models to gaming, animation, and interactive experiences. MIXI reports significant production efficiencies, including reducing a 21-day video workflow to three days and cutting visual proposal creation time from several days to five minutes.
RunwayJun 30Runway now offers Seed Audio 1.0 to all paid plans, enabling the generation of up to 120 seconds of speech, sound design, and music from text prompts. The model supports multi-speaker dialogue and voice cloning through uploaded audio reference snippets. This integration provides tools for creating immersive soundscapes and custom audio content directly within the Runway platform.
RunwayJun 25Runway launched Agent 2.0, an autonomous creative partner that generates marketing briefs and campaign assets from simple prompts. The system analyzes performance data to optimize creative and scales content across platforms, formats, and markets. Agent 2.0 is available now on the Runway web platform to handle end-to-end marketing workflows.
RunwayJun 25Runway now hosts third-party video generation models, including Seedance 4K, Seedance Mini, and Kling 3.0 Turbo, directly on its platform. This expansion provides access to multiple industry-leading models within a single interface.
RunwayJun 24Runway launched an ad localization tool that generates market-ready versions of advertisements from a single image. The system translates ad copy into over 20 languages while preserving the original visual layout. This one-click workflow automates the adaptation of static creative assets for international markets.
RunwayJun 18Runway launched Recipes, a set of pre-built API endpoints that package the company’s prompting and workflow expertise. These endpoints integrate production-ready generative media features, such as automated product ad generation and video product swapping, with a single API call. Recipes eliminate the need for external platforms to build or maintain complex generative workflows.
RunwayJun 17Runway launched an official ChatGPT App that generates and edits video and images directly within ChatGPT conversations. The integration accepts prompts, images, and product URLs to create marketing spots, social clips, and creative content. This app removes the need for tab-switching, allowing for end-to-end media creation inside the chat interface.
Frequently asked questions