Stensyl Write Studio: AI Drafting for Creative Projects.

What Write Studio Actually Is

Write Studio is Stensyl's long-form document drafting surface, available at /write. It is where briefs get written, scripts take shape, proposals move from rough outline to client-ready copy, and campaign narratives find their structure. It is not a chat window. It is not a caption generator. It is a dedicated drafting environment built for the kind of writing that takes more than three sentences to do properly.

That distinction matters. Most AI writing tools present as either a chat box or a template gallery. You type a prompt, you get a paragraph, you copy it elsewhere and carry on. Write Studio is closer in philosophy to a document editor that happens to have six frontier language models sitting inside it. You work in it. You build in it. The document stays open while you think, revise, and switch models.

The model picker inside Write gives you six options: GPT-5.4 mini, Gemini Flash, Gemini Pro, GPT-5.5, Claude Sonnet 4.6, and Claude Opus 4.7. All six are available on every plan, including Free. Model access is never a tier decision on Stensyl. The only things that change with your plan are how many credits you have each month and how many simultaneous generations you can run.

The same six models are also accessible via the Canvas LLM Chat node at /canvas, where written outputs can pipe directly into image generation nodes, video nodes, or any other downstream step in a node-based workflow. Write Studio is the long-form surface; Canvas is where that writing connects to the rest of a generative pipeline. Both pull from the same model set.

One clarification worth making early: there is no standalone surface on Stensyl called "Chat mode." Multi-model text generation lives in Write for structured, long-form work and in the Canvas LLM Chat node for workflow chaining. That is the full picture.

Six Models, One Surface: Choosing the Right Tool

Having six models available sounds generous until you realise the choice is meaningful. These are not six versions of the same output. Each model has a different balance of speed, depth, tone sensitivity, and cost-in-credits. Matching the model to the task is one of the faster ways to improve both the quality of what you get and how efficiently you spend your credits.

Fast, High-Volume Drafting

GPT-5.4 mini and Gemini Flash are the workhorses for high-volume, shorter-form output. Think social captions across ten product variants, a batch of email subject lines for an A/B test, ad copy variations for a paid campaign, or UX microcopy drafts for a component library. These models are fast and credit-efficient. A marketing team generating thirty headline variants for a campaign launch, or a product designer writing descriptions for an entire product catalogue, will reach for one of these first.

The trade-off is nuance. On a 300-word product page, GPT-5.4 mini will give you something competent and fast. On a 2,000-word brand narrative with careful tonal requirements, it will start to feel thin.

Structured, Multi-Section Documents

GPT-5.5 and Gemini Pro handle longer, more architecturally complex documents, where consistency across sections matters as much as the quality of any individual paragraph. An exhibition design brief covering spatial concept, visitor journey, material palette, and lighting rationale needs a model that holds its argument across all four sections. A game design document with mechanics, narrative framework, and progression systems is the same challenge. These models are better suited to it.

Gemini Pro is particularly strong on structured documents where logical coherence is the priority. A UX copy framework covering onboarding flows, error states, empty states, and help text needs to feel like one writer produced it. Gemini Pro handles that continuity well.

Tone-Sensitive and Narrative Work

Claude Sonnet 4.6 and Claude Opus 4.7 from Anthropic are widely regarded across the industry for their sensitivity to voice and their ability to follow complex stylistic instructions without drifting. Claude Sonnet 4.6 is the balanced choice: articulate, responsive to brief-level constraints, and strong for brand voice guidelines, pitch decks, and scripts where tone carries the argument.

Claude Opus 4.7 is where you go when voice and depth matter more than speed. A film treatment where atmosphere has to land on the page. A complex client proposal for an interior design project where the writing needs to feel considered, not generated. A character bible for a game where every NPC needs a distinct but internally consistent voice. Opus 4.7 produces writing that reads like someone thought carefully before committing to a sentence. It costs more credits. For the right document, it is worth it.

Model	Best for	Relative speed
GPT-5.4 mini	Social captions, product descriptions, ad copy variants, SEO meta	Fast
Gemini Flash	Email subject lines, microcopy drafts, short-form batch work	Fast
Gemini Pro	UX copy frameworks, IA-linked copy, structured brand documents	Moderate
GPT-5.5	Exhibition narratives, game design docs, multi-section proposals	Moderate
Claude Sonnet 4.6	Brand voice guidelines, pitch decks, scripts, explainer copy	Moderate
Claude Opus 4.7	Film treatments, character bibles, complex client proposals	Deliberate

Practical Use Cases Across Creative Disciplines

Write Studio does not have a natural home in any one discipline. The following are concrete examples of how it works across the twelve disciplines Stensyl serves.

Game Development

Game writers and narrative designers use Write Studio to draft quest outlines, NPC dialogue trees, item descriptions, and player onboarding copy in a single session. The standard approach: use GPT-5.5 or Claude Opus 4.7 to produce a narrative design document with world context, faction logic, and main quest structure, then switch to GPT-5.4 mini to generate batches of item descriptions and incidental dialogue variations at volume. The longer model handles lore consistency; the faster model handles throughput. Both live in the same document surface, so context is never lost between passes.

Marketing and Advertising

Agencies use Write Studio to move from campaign brief to copy assets without jumping tools. A typical session: draft the campaign strategy document in Claude Sonnet 4.6, use the same session to generate headline pools and tagline variants in Gemini Flash, then carry the strongest lines into Marketing Studio at /marketing-studio for formatting into carousels, ads, and post formats. The drafting surface is where thinking happens; Marketing Studio is where output goes to market.

Teams building brand-voice-trained workflows can paste a style guide directly into the prompt context, giving any of the six models a reference point to work from. This is standard practice in agency copywriting and Write Studio supports it without any additional configuration.

Interior Design and Exhibition Design

Designers who are visually fluent but time-poor on long-form narrative use Write Studio to turn concept sketches and moodboards into client-facing copy. An exhibition design team pitching a spatial concept needs more than floor plans: they need a narrative that explains visitor journey, material choices, and the emotional experience of the space. Claude Opus 4.7 handles this well. The same workflow applies to interior design practices writing material rationale documents or spatial storytelling sections for pitch decks.

Motion Design and Film

Motion designers and film directors use Write Studio to draft treatments, voiceover scripts, beat sheets, and director's notes before moving into storyboard and animatic work. The practical benefit is structure: having a scene-by-scene treatment written out in Write gives you a ready-made prompt library for downstream work in Film at /film or in Canvas video nodes. The written document and the visual pipeline stay in the same project, so the copy and the frames reference the same intent.

Web and UX Design

Product and UX teams face a specific problem: wireframes move fast, and copy almost always follows weeks behind. Write Studio changes that ratio. A UX writer can draft microcopy, onboarding flows, error states, and landing page copy at wireframing speed, using Gemini Pro for structured copy frameworks or GPT-5.4 mini for rapid iterations on specific components. The output drops directly into the design system rather than arriving as a late-stage handoff.

Copy must keep up with wireframing speed, not follow weeks later. Write Studio gives UX teams a drafting surface that works at the pace of the design process, not the pace of a traditional copywriting brief.

Write Studio Inside a Wider Stensyl Workflow

Write Studio gains most of its value when it sits inside a connected project rather than as an isolated tool. The way Stensyl is structured, every surface shares a project layer at /projects, where brand identity, assets, and team workspaces are organised. A document drafted in Write belongs to a project that also holds the image outputs, video renders, and 3D assets generated from the same brief. That connection is the difference between a useful writing tool and an integrated creative workflow.

Research to Writing

The Research surface at /research is Perplexity-backed and designed for competitive scans, reference gathering, and market context. The standard workflow: pull relevant references and competitor positioning in Research, then move directly into Write to draft a brief or campaign document against those findings. No tool-switching. No copy-pasting between browser tabs. The research and the writing live in the same environment.

Writing to Visual Output

A scene treatment drafted in Write can become the prompt for a Film session at /film. A brand narrative can inform the style prompt for image generation at /generate/image. A product description drafted in Write can feed into a social carousel in Marketing Studio. The written document does not end at the edge of the drafting surface; it becomes the brief that drives visual output.

Canvas: Writing in a Node Workflow

The Canvas LLM Chat node at /canvas extends the same six writing models into a node-based pipeline. A copywriting output, say a batch of ad hooks, can pipe directly into an image generation node that uses each hook as a prompt. A voiceover script can connect to an Audio node for text-to-speech rendering. This is where the model-agnostic approach becomes genuinely structural: one model call producing written output that branches into three different generative streams simultaneously. The creative brief stops being a document someone reads and starts being a live input to the pipeline.

Credits, Plans, and Getting Started

Write Studio is available on every plan, including the Free tier. Free accounts receive 150 one-time credits and access to all six writing models from day one. No credit card required. Those credits do not reset, but they are enough to run a meaningful first session and compare models before committing to a plan.

From there, paid plans scale on credits and concurrency:

• Lite (£10/mo): 1,000 credits per month, 1 concurrent generation, 20 daily chat-widget messages.
• Starter (£22/mo): 2,500 credits per month, 2 concurrent generations, unlimited chat-widget messages.
• Pro (£42/mo): 6,000 credits per month, 3 concurrent generations, unlimited chat-widget messages.
• Studio (£84/mo): 12,500 credits per month, 4 concurrent generations, unlimited chat-widget messages.

The concurrency limit applies across the whole platform, not just Write. A Pro user running an image generation and a video render simultaneously has one remaining concurrent slot. For teams running heavy multi-surface workflows, where Write output is feeding Canvas pipelines that are simultaneously triggering image and video nodes, the Studio tier's four concurrent generations gives the most headroom.

Context for the pricing: most dedicated AI copywriting tools charge £10 to £20 per month for a single underlying model, with volume limits on output. They do not expose multiple frontier models. Stensyl's model-agnostic approach means a Free-tier user has access to GPT-5.5, Claude Opus 4.7, and Gemini Pro alongside the lighter models. The choice between models is always a creative decision, never a billing one.

Where to Start

For new users, the most useful first session is also the most instructive one. Open Write, select Claude Opus 4.7, and draft a brief for your current project: a campaign, a design concept narrative, a game narrative document, whatever is live right now. Then run the same prompt in Gemini Flash. The difference in output, not just quality but character, gives you a working sense of where each model sits on the speed-versus-depth axis. That comparison is more useful than any documentation.

From there, the workflow tends to find its own shape. Opus for the documents that need to hold an argument. Flash for the outputs that need volume. Sonnet for the copy where tone is the deliverable. GPT-5.5 for the brief that has eight sections and needs to read like one person wrote it. The model picker is not a preference setting. It is a creative decision, and Write Studio is where you make it.