Skip to content

How to document Runway work for client delivery

Runway organizes video generation into sessions and assets, which is good working storage and a thin delivery record. The settings that produced a shot live in the browser tab, not in the file you ship.

Published July 22, 2026.

Runway sits in more production pipelines than any other video generation tool, and its interface is built around iteration: you prompt, you watch takes accumulate in a session, you pick one, and you move on. The help center describes sessions as the place where you create generations and as a folder or group of generations made at a given time, with outputs saving automatically to the Assets panel. Completed generations can be viewed, downloaded, and managed through the asset browser, and Private is the default storage location. As working storage during a project, this is well organized and easy to navigate.

As a record of how a delivered shot was made, it is thinner than it looks. The export is a plain video file. The text prompt, the input image or keyframes, the model version, the duration and resolution settings, and the seed all stay in the session view. Runway's Gen-4 documentation is explicit that a Fixed seed produces generations with similar motion rather than identical ones, so a seed is a way to get sibling takes, not a replay button. And because video work produces many near-identical candidates per shot, the question that matters most at delivery time, namely which take the client approved and when, is not a concept the tool has at all.

The practical consequence is that Runway documentation happens at two moments: when you generate, and when the client approves. At generation time you capture the inputs and settings while the session is open in front of you. At approval time you pin the exact take, export it, and connect it to the sign-off. Both are quick while the work is fresh, and close to impossible to reconstruct honestly months later.

What Runway stores on its side

  • Runway organizes work into sessions, described in the help center as where you create generations and as a folder or group of generations you create at a given time. Source
  • Completed generations can be viewed, downloaded, and managed through the asset browser, and Private is the default asset storage location alongside Shared and Favorited. Source
  • Gen-4 video supports a Fixed seed setting, and Runway documents that you can copy and paste the seed of a previous output to create generations with similar motion. Source
  • Gen-4 renders takes at five or ten seconds and offers six output resolutions, from vertical 9:16 at 720x1280 up to widescreen 21:9 at 1584x672. Source

What gets lost without your own record

  • The exported video is a plain file. The prompt, the input image, the seed, the model version, and the duration and resolution settings stay in the Runway session and do not travel with the download.
  • The help center does not document any export of the generation record itself, so the session view inside your account is the only place those settings exist.
  • A Fixed seed gives you similar motion, in Runway's own words, not an identical regeneration, so the approved export is the only reliable copy of the delivered shot.
  • Assets default to private storage, so a take generated under a freelancer's account sits outside the team's reach unless it was deliberately shared to the workspace.
  • Nothing in a session records which take the client approved, which were rejected, or why, and with dozens of near-identical candidates per shot that distinction is the record that matters.

The documentation checklist for Runway

  1. Record the full text prompt and archive the exact input image or keyframes for every kept take, because image-to-video results cannot be interpreted without their inputs.
  2. Note the model and version that produced the take, for example Gen-4 or Gen-4.5, since Runway ships new models often and the same prompt behaves differently on each.
  3. Enable Fixed seed when a shot may need sibling takes later, and write the seed down next to the prompt rather than leaving it in the session.
  4. Record the duration and the resolution setting per take, since both are choices you will be asked about.
  5. Name sessions after project and shot so the structure inside Runway mirrors your delivery structure, and share client work to the workspace instead of leaving it in private assets.
  6. At sign-off, export the approved take immediately, record its generation date and session, and connect the file to the approval instead of trusting the asset browser as the archive.

Generated video sits squarely inside the transparency obligations of Article 50 of the EU AI Act, which reaches synthetic media shown to the public, with penalties for transparency violations of up to €15 million or 3% of worldwide turnover. Runway's exports carry no disclosure for you, so what was generated, how it was labeled, and when it went out is your record to produce.

Keep the claim structure of that record honest. It can establish that a prompt, an input image, a seed, and an approved export existed on a date and have not changed since, which is the notarial layer a client dispute actually needs. It cannot establish that a particular input caused a particular frame, and it does not need to. A producer who can put the approved take, its settings, and its sign-off date on the table has answered every question that realistically comes back.

Frequently asked questions

Can I recover the prompt and settings of an old Runway generation?

While the generation still exists in your account, yes: open the session or find the asset in the browser and read the settings from the generation view. None of that is inside the video you exported, so if the account or the session goes, the settings go with it.

If I saved the seed, can I regenerate the exact same shot later?

No. Runway's documentation describes a Fixed seed as producing generations with similar motion, which is useful for sibling takes and continuity, but it is not a replay of the approved shot. Keep the approved export itself as the reference copy.

Where do my Runway generations actually live?

They accumulate in the session where you created them and save automatically to the Assets panel, where you can view, download, and manage them. Private is the default storage location, so shared team access only exists where someone chose to share.

Does the exported video disclose that it was AI generated?

The help center does not document any embedded generation record or disclosure in the download, so plan your EU AI Act Article 50 labeling and your production record independently of the file.

What should the record say when a client signs off a Runway shot?

Which take was approved, from which session, with its prompt, input image, seed, model version, duration and resolution, plus the export date and the name of the approver. That is one short entry if you write it at sign-off and a reconstruction project if you write it later.

Related guides

  • How to document AI-generated work for client delivery What broadcasters, platforms, agencies, and rights offices already require when AI is part of delivered work, and the per-shot checklist that answers all of them.
  • EU AI Act Article 50 for creative teams What the transparency obligations for AI-generated and AI-manipulated content actually say, the real timeline and penalty ceiling, and what a team should record to be ready.
  • AI provenance: what you can prove and what you cannot An honest map of provenance claims: what a record can demonstrate about inclusion, integrity, and time, and which causal claims no system can support.
  • Prompt and model documentation: a practical checklist for studios The fields worth capturing for every generation: prompts, models, versions, settings, references, and edits, and how to keep the habit alive in production.
  • What is an AI production log? An AI production log is the per-asset record of prompts, models, settings, references, edits, rights decisions, and approvals behind AI-assisted work. Here is the minimum useful structure and the limits of what that record can prove.
  • C2PA vs. AI workflow documentation C2PA and AI workflow documentation solve different parts of the provenance problem. C2PA protects signed assertions attached to media; workflow records preserve prompts, sources, decisions, rights, versions, and approvals around the file.
  • AI shot log template for image and video production A practical AI shot log template for prompts, models, references, versions, human edits, disclosure, and approval. Download the CSV and adapt the field guide to your production workflow.
  • How to record client approval for AI-generated work A client approval record for AI-generated work should identify the exact version, disclosed AI use, review scope, requested changes, approver, and time. This guide turns sign-off into evidence instead of an ambiguous email.
  • Guide editorial and verification policy How Behind The Workflow researches, sources, dates, updates, and corrects its AI production guides. Product facts use primary sources, legal claims link to official text, and machine-readable markup mirrors visible content.
  • How to document Sora work for client delivery Sora embeds C2PA metadata in every asset and marks downloads with a visible moving watermark, but the prompts, iterations, and cameo consents behind a clip stay in the app. This guide covers what OpenAI's provenance signals prove and what you still have to record before client delivery.
  • How to document Google Veo work for client delivery Every Flow output made with Veo carries an invisible SynthID watermark, and Gemini can verify it, but the check names Google AI as the origin and nothing else. This guide covers what to record so a Veo delivery can answer the questions SynthID cannot.
  • How to document Luma Dream Machine work for client delivery Dream Machine keeps your work organized in Boards and an Ideas stream, but the watermark and the commercial license of a clip depend on the plan that generated it, and the exported file names neither. This guide covers the records that close that gap before client delivery.
  • How to document Ideogram work for client delivery Ideogram shows the seed and even the Magic Prompt rewrite for every image in its Details Panel, but none of it travels with the downloaded file, and images are public by default. This guide covers what to copy out of the panel and when to generate privately.
  • How to document Midjourney work for client delivery Midjourney keeps a searchable archive of everything you generate, but the record stays inside your account and images are publicly discoverable by default. This guide covers what midjourney.com actually stores, what never travels with a delivered file, and the per-job checklist that closes the gap.
  • How to document ComfyUI work for client delivery ComfyUI automatically embeds the workflow graph in the metadata of every image it saves, which makes it unusually self-documenting. This guide covers what that embedded workflow proves, where it silently disappears, and what a client record still needs on top.
  • How to document self-hosted Stable Diffusion work for client delivery A self-hosted Stable Diffusion setup writes its generation parameters into every image by default, at least in the AUTOMATIC1111 web UI most teams run. This guide covers what that buys you, what silently breaks it, and what a client record still needs.
  • How to document FLUX work for client delivery The FLUX API from Black Forest Labs delivers each image through a signed URL that stays valid for ten minutes, and the docs describe no archive to return to later. This guide covers what the API gives you at request time and the logging your pipeline must do in that window.
  • How to document Adobe Firefly work for client delivery Firefly attaches a Content Credential naming the model and version to what it generates, and its API returns the seed for every image. This guide covers what that provenance layer proves, how it can be stripped, and what a client delivery record still adds.

Every requirement in this guide is answered from a record kept while the work happens. BTW keeps that record for you.

Start your own record