From Brief to Deliverable: Stensyl's Write Studio Explained.

Why Tab-Switching Breaks Creative Momentum

The typical creative workflow is not one tool. It is five tools wearing the costume of a process. A brief starts in Google Docs. Research opens in a second browser tab, then a third. The AI writing tool is a fourth app entirely. By the time a motion designer is refining storyboard copy mid-render, or a marketing strategist is cross-referencing a campaign brief with a social copy doc, the original problem they sat down to solve has quietly dissolved into the friction of moving between surfaces.

This is not a minor inconvenience. Research on workplace attention, most notably Gloria Mark's work on interruption and recovery, consistently finds that knowledge workers take anywhere from 9 to 25 minutes to fully resume focused work after an interruption. For creatives, the damage is not just lost minutes. It is lost context. Every time you paste a paragraph from Notion into a separate AI tool, you are re-explaining the brief from scratch. The model you just opened has no idea what you were solving for. You do, but faintly.

Consider a few common examples. A UX designer finishes a wireframe rationale in Figma, opens a separate tab to refine the copy in ChatGPT, then pastes the result back into a Notion page that is already three versions behind the Figma file. A motion designer writes a storyboard brief, pauses to generate a test render, then returns to the brief to discover they have lost the thread of the visual direction they were building. A marketing strategist bounces between a campaign positioning document and a social copy brief, updating both independently until neither reflects the other accurately.

In each case, the problem is not a lack of good tools. It is a lack of continuity between them. The handoff between apps introduces version drift, re-explanation overhead, and a slow erosion of the original intention behind the work.

Stensyl's Write studio, Research surface, and Canvas node-based editor are designed around a different assumption: that drafting, refining, and generating should happen in a single environment, connected by context rather than separated by copy-paste.

What the Write Studio Actually Does

Write is Stensyl's long-form document drafting surface. It is not a chat window. It is not a notes app. It is purpose-built for the kind of documents creative professionals actually produce and need to version, refine, and connect to other outputs.

That means project briefs, copy decks, design rationales, campaign frameworks, exhibition scripts, game narrative outlines, brand identity documents, and concept presentation write-ups. Documents with length, structure, and revision cycles. Documents that need to end up somewhere useful, not just in a downloads folder.

The defining feature of Write is the model picker. Six writing models are available on every plan:

• GPT-5.4 mini (OpenAI)
• GPT-5.5 (OpenAI)
• Gemini Flash (Google)
• Gemini Pro (Google)
• Claude Sonnet 4.6 (Anthropic)
• Claude Opus 4.7 (Anthropic)

This is not a token of flexibility. It changes how you approach a document. You choose the model for the task at hand, rather than accepting whatever default the platform decided was good enough.

Gemini Flash is fast and responsive, well suited to rapid first-draft iteration when you need volume on the page quickly. Claude Opus 4.7 is the better choice for documents that require structured reasoning and nuanced tone control: a tone-of-voice guide, a complex exhibition brief, a brand strategy document where the argument needs to hold together across ten sections. GPT-5.5 earns its place when the output will feed into something technical or structured, a content schema, a detailed product specification, a UX content framework with clear taxonomies.

It is also worth being clear about what Write is not. Multi-model chat that pipes into generative outputs lives in the Canvas LLM Chat node, which is part of Stensyl's node-based workflow editor. Write is for documents. Canvas is for connected workflows. Both surfaces share the same six-model picker, but they serve different stages of work.

Projects connect Write to everything else. A brief drafted in Write sits in the same project workspace as Generate outputs, Moodboards, Storyboards, and Social assets. It is not a floating document. It has a home, and that home is the same place as the work it informed.

Feeding Research Directly Into Your Draft

The Research surface on Stensyl is backed by Perplexity, which combines web search with a language model to return cited answers rather than raw links. For creative professionals, this matters most when a document needs factual grounding that goes beyond what you already know.

A product designer researching material sustainability benchmarks for a brief does not want to open five browser tabs, skim three articles, copy fragments into a separate notes doc, and then open their drafting tool and re-explain all of it. An exhibition designer scoping venue specifications for a stand concept faces the same problem. The information exists. The overhead is in retrieving it and connecting it to the work.

The practical workflow inside Stensyl runs like this: open Research, query for the factual or contextual grounding you need, and move directly to Write with that reference sitting in the same platform session. No app switch. No losing your place. No re-explaining to an AI tool that has never seen the document you are building toward.

Compare this to the alternative workflow most creatives are running today: open Google, scan results, open three pages, copy key points into a separate notes document, open the AI writing tool, paste the brief, paste the notes, re-contextualise everything, and start drafting. That process is not broken exactly. It is just slow, and every step where context passes through your clipboard is a step where something gets lost or left behind.

Ray, Stensyl's creative-decision assistant, fits into this flow as a routing tool rather than a drafting tool. If you are unsure whether a particular document type belongs in Write, whether it needs Research input first, or which of the six writing models is a sensible starting point, Ray can orient you quickly. It is locked to a fast Anthropic Haiku model and is designed for speed, not depth. Think of it as a well-informed colleague who knows the platform well, not as a co-writer.

The shift from research to drafting should not require switching tools. Keeping both in the same environment is not a convenience feature. It is how you keep the document coherent from the first line.

Switching Models Mid-Document Without Starting Over

The model picker in Write is not a one-time setup. You can switch models at any point in a drafting session. This is a practical capability, not a headline feature, and it becomes genuinely useful the moment you are working on a document that has different demands across its sections.

Consider a graphic designer drafting a brand identity proposal. The strategic positioning section needs to be generated quickly, covering the core territory before refining it. Gemini Flash handles that fast. The tone-of-voice guidelines, which need precision, careful reasoning, and real sensitivity to how language lands, are a different kind of task. Claude Opus 4.7 is the better model there. Both sections live in the same document. The model switch takes seconds.

Or consider an automotive designer building a concept presentation brief. The technical specification sections, covering powertrain references, dimension parameters, and materials callouts, benefit from GPT-5.5's structured output style. The narrative framing of the visual direction, where you are trying to convey mood and intention rather than data, suits Gemini Pro's more expansive prose. One document, two models, no second subscription and no version drift between apps.

This is the practical shape of what industry practitioners increasingly describe as the "fast + cheap, then heavy" pattern. Use lighter, faster models for first-pass drafting and bulk generation. Reserve the more capable, credit-intensive models for the sections that genuinely need them: the argument that has to land, the rationale a client will read closely, the copy that represents the brand.

The credit system on Stensyl is unified across all six models on every plan. Every model draws from the same credit pool. This matters because it removes the incentive to use only one model just to avoid managing multiple billing systems. It also means you should be deliberate about where you apply heavy models, because Claude Opus 4.7 costs more credits per generation than Gemini Flash. On a long document, using Opus for every section is not necessary and not efficient. Using it for the three sections that need it is.

Model	Best used for	Credit weight
Gemini Flash	Rapid first drafts, high-volume iteration, fast structural outlines	Light
GPT-5.4 mini	Efficient drafting, content variants, structured list generation	Light
Gemini Pro	Narrative sections, translation, richer reasoning within a draft	Mid
GPT-5.5	Technical specifications, structured deliverables, content schemas	Mid
Claude Sonnet 4.6	Editing passes, tone refinement, client-facing copy	Mid-heavy
Claude Opus 4.7	Complex reasoning, nuanced tone-of-voice, high-stakes document sections	Heavy

From Written Brief to Visual Output via Canvas

Write produces documents. Canvas makes those documents generative inputs.

Canvas is Stensyl's node-based workflow editor. Inside it, the LLM Chat node provides multi-model chat that can pipe directly into other generation nodes, including Image Generate and Video nodes. The connection between a written brief and a visual output is not conceptual on Stensyl. It is a literal node connection in a workflow.

A film set designer who has drafted a scene brief in Write can move that brief into a Canvas workflow, run it through the LLM Chat node to extract the key visual prompts from each scene description, and feed those prompts directly into the Generate surface without manual copy-paste. The brief does not become a reference document. It becomes an active input in the pipeline.

For game development, the workflow is equally direct. A narrative designer drafts environment lore in Write: the atmospheric detail of a collapsed industrial zone, the material logic of a hand-built settlement, the colour language of a faction's territory. In Canvas, they pass selected passages through the LLM Chat node to extract image generation prompts. Those prompts connect to image generation nodes and produce concept art reference. The lore document and the visual output are part of the same workflow, not separate files in separate tools.

Exhibition design offers another clear example. A stand concept brief drafted in Write, covering the spatial narrative, material palette, and key visitor experience moments, can inform a Moodboard via the project workspace. Simultaneously, a Canvas workflow uses that same brief to generate spatial visualisation prompts and feeds them into Generate outputs. The brief is never duplicated. It is referenced in place.

The node-based structure of Canvas changes the status of a written brief from a static document to a live input. That distinction matters practically. When the brief changes, the workflow can run again. The outputs update. The intention stays intact.

Organising Multi-Stage Work Inside Projects

Projects on Stensyl are shared workspaces that hold brand identity, team context, and all creative outputs in one place. A brief drafted in Write does not float loose in the platform. It lives in the project alongside Generate outputs, Storyboards, Moodboards, and Social assets. Everything connected to a piece of work occupies the same space.

For teams, this changes the coordination dynamic in a practical way. A junior designer and a creative director can work from the same brief without emailing document versions or maintaining parallel Notion pages. The project workspace is the single source of truth. When the brief changes, everyone working in that project sees the current version. There is no question of which document to open.

The clearest example of this for marketing and advertising teams: a campaign brief drafted in Write, visual concepts generated via Generate, and social carousel assets produced in the Social studio all sit in the same project. The brief informs the visual direction. The visual outputs reference back to the brief. When the client asks for the rationale behind a particular image choice, the answer is one click away rather than buried in an email thread or a Slack message from three weeks ago.

It is worth being precise about what Projects do not do. They are not a project management system with timelines, task assignment, or approval workflows. They are an organisational layer. Their purpose is to connect creative assets and keep context intact across a multi-stage workflow. If your team needs full project management, that sits outside Stensyl. What Projects solve is the specific problem of assets and documents drifting apart as work progresses.

Pricing context is relevant here. Stensyl is available from £10 per month on the Lite plan, which includes 1,000 credits and access to all six writing models. The Starter plan at £22 per month extends that to 2,500 credits and removes the daily chat message limit. Pro at £42 per month and Studio at £84 per month scale credits further for heavier workloads. Every model and every surface is open on every plan. Tiers differ by how many credits you have available, not by what you can access.

The brief is not just a starting point. It is the thread that should run through every output. Projects on Stensyl keep that thread intact from the first sentence to the final exported asset.

The shift from brief to deliverable is not about moving faster, though that is a real benefit. It is about keeping the original intention legible at every stage of the work. When the brief lives in a separate app from the research, which lives in a separate app from the drafting tool, which lives in a separate app from the generation surface, something gets lost at each boundary. Not catastrophically. Just incrementally, and then persistently.

Write, Research, Canvas, and Projects are the architecture Stensyl has built to close those gaps. Draft the brief with the model that suits the document. Pull research without opening a browser. Switch models mid-document when the section demands it. Move the brief into Canvas and make it a live input to a generative pipeline. Keep everything in a project so the work holds together from the first line to the last export. That is the complete path from brief to deliverable, and it runs inside one platform.