Runway ML

Learning Objectives

• Understand what Runway ML is and how it differs from pure text-to-video generators
• Learn the core tools Runway offers beyond video generation
• Identify the use cases where Runway's creative-professional focus adds the most value

What Is Runway ML?

Runway ML (commonly called Runway) is a New York-based AI creative platform founded in 2018. Unlike Veo and other text-to-video generators embedded in larger AI ecosystems, Runway is purpose-built as a professional creative tool — combining AI video generation with editing, compositing, and media production capabilities in a single web-based workspace.

Runway gained widespread recognition when its Gen-1 and Gen-2 models were used by professional filmmakers and in Hollywood productions. Its current flagship, Gen-4.5, released December 1, 2025, took the top position on the independent Artificial Analysis text-to-video benchmark — and unusually for a quality jump, Runway says it holds the speed and efficiency of Gen-4 rather than trading against them.

Pricing

Paid prices above are the standard monthly rate; billing annually cuts each to roughly $12, $28, and $76. Runway renamed its tiers, and the old names now point at different plans — what was Basic is now Standard, the old Standard is now Pro, and the old Pro is now Max, so a comparison written against the previous naming will be wrong on both price and allowance.

Credits are consumed per generation, and Runway publishes the exchange rate directly: 625 credits buys about 52 seconds of Gen-4.5 or 104 seconds of the faster Gen-4 Turbo. Length is governed by your credit balance rather than a hard per-tier clip cap, so the practical question is how many seconds per month you need, not which tier unlocks longer output.

Core Capabilities

Gen-4.5 — Text-to-Video

Runway's flagship generation model, and the one to reach for when output quality is the point:

• Benchmark position: ranked first on the independent Artificial Analysis text-to-video leaderboard at launch
• Camera control: explicit camera motion instructions (dolly, crane, orbit, tilt)
• Temporal consistency: strong object and scene persistence across frames, with improved control over dynamic action
• Efficiency: roughly the speed and cost profile of Gen-4, which is what makes it a default rather than a special-occasion model
• Availability: on every paid plan, not gated to the top tier

Image-to-Video

One of Runway's most used features: upload a static image and extend it into motion video. A product photograph becomes a rotating 3D showcase. A landscape photo becomes a slowly moving aerial scene. Character art becomes an animated sequence.

Video-to-Video (Style Transfer)

Apply AI visual styles to existing video footage — transform a standard recording into a cinematic style, animated look, or specific aesthetic without re-shooting.

Gen-4 Turbo — Fast Iteration

The speed-first model, and roughly twice as credit-efficient as the flagship: the same 625 credits buys about 104 seconds of Gen-4 Turbo against 52 seconds of Gen-4.5. Use it during ideation, when seeing many options quickly matters more than final quality, then re-run the shot you picked on Gen-4.5.

Aleph — In-Context Video Editing

Released July 2025, Aleph edits footage you already have rather than generating from nothing. It reads a scene's structure, lighting, depth and motion, then applies instructions expressed in plain language: remove or replace an object, change the season or weather, relight the shot, add visual effects, or generate a new camera angle on the same scene. For post-production teams this is often more useful than text-to-video, because the input is real footage.

Act-Two — Performance Capture

Drives a generated character from a recorded performance, transferring an actor's motion and expression rather than describing them in a prompt.

Additional Creative Tools

Runway offers a suite of tools beyond video generation:

• Inpainting — erase objects from video frames and fill them with AI-generated content
• Green Screen (Background Removal) — automated background removal for video without a physical green screen
• Frame Interpolation — smooth slow-motion effects from standard frame-rate footage
• Super Slow-Mo — AI-enhanced frame rate increase for dramatic slow motion
• Text-to-Image — image generation alongside the video tools

Strengths

• Benchmark-leading generation — Gen-4.5 took the top spot on Artificial Analysis' text-to-video ranking, and did it without giving up Gen-4's speed
• Professional creative platform — built for filmmakers, designers, and studios; not just a generator but a workspace
• Industry adoption — used in actual film and TV production; strong credibility with professional creative teams
• Editing, not just generating — Aleph works on footage you already shot, which is the workflow most production teams actually need
• A real model ladder — Gen-4.5 for finals, Gen-4 Turbo at roughly half the credit cost for iteration, Aleph and Act-Two for editing and performance
• Suite of tools — background removal, inpainting, interpolation alongside generation in one platform
• Strong community — active creator community with tutorials, templates, and shared generations for inspiration

Limitations & Considerations

• Credit system — credits run out, and the entry tier's 625 credits is about 52 seconds of flagship output per month; heavy use means the Max tier or top-ups, which gets expensive at volume
• Tier names were reused — Runway's Standard, Pro and Max are not the Basic, Standard and Pro of its earlier pricing, so older comparisons and blog posts describe the wrong plan at the wrong price
• Short clips — generations are measured in seconds, not minutes; Runway is a shot-generation tool, not long-form video
• Text rendering — AI video models including Runway do not reliably render readable text inside video frames
• Privacy: Generations on free and lower tiers may be visible in Runway's community gallery by default; disable under account settings or use Teams/Enterprise for private workflows

Best Use Cases

When to choose alternatives:

• Fastest text-to-video with audio → Veo 3 or Kling AI
• Avatar-based presenter video → HeyGen or Synthesia
• Completely free experimental use → Pika Labs
• AI video editing (transcription, voice editing) → Descript

Getting Started

1. Go to runway.com and create a free account — 125 one-time credits to start
2. Click New Project → select Video Generation → choose Gen-4.5 for quality or Gen-4 Turbo for speed
3. Choose Text to Video or Image to Video
4. For text-to-video: write a detailed prompt including subject, camera motion, lighting, and style
5. For image-to-video: upload your image, add a motion prompt describing how it should move
6. Iterate on Gen-4 Turbo at roughly half the credit cost, then re-run your chosen shot on Gen-4.5 for the final
7. To edit footage you already have rather than generate new video, start from Aleph instead

Key Takeaways

• Runway ML is a professional AI video platform — not just a generator, but a creative workspace combining generation, editing, and compositing tools
• Gen-4.5, released December 2025, is the current flagship and took first place on the independent Artificial Analysis text-to-video benchmark while keeping Gen-4's speed
• The lineup is a ladder, not a single model: Gen-4 Turbo for iteration at roughly half the credit cost, Aleph for editing footage you already shot, Act-Two for performance capture
• Image-to-video is a standout feature: animate your existing visual assets into motion without starting from scratch
• Check the tier name against the price before comparing plans — Runway's Standard, Pro and Max replaced an earlier Basic, Standard and Pro, so the same words now mean different plans
• The credit system means Runway suits creators who need quality over volume; for high-throughput generation, Veo 3 or Kling AI offer better scalability