
Deepest reasoning. Best for complex, multi-page documents.
Anthropic's flagship model. Claude Opus 4.6 excels at chained logic, multi-step deductions, and cross-referencing long documents. For a 15-page planning statement that must align site constraints, policy documents, and design rationale without contradiction, Opus is measurably more reliable than any other model. 1M token context window.
Type what you need: a proposal, case study, specification, or creative copy.
Pick your writing model and document length. See the credit cost before you generate.
Your document is drafted and ready to review, edit, copy, or export. Iterate or switch models.
Jump into the Studio and start writing. Plans from £10/month.
Choose a PlanClaude Opus 4.6 is Anthropic's most capable reasoning model. It processes prompts through a multi-step chain of thought, cross-referencing context, checking for contradictions, and building arguments with the kind of structured logic that professional documents demand. For design professionals producing planning statements, specifications, or detailed proposals, this depth of reasoning is not optional.
The 1M token context window means Opus can hold an entire project brief, site constraints document, relevant policy extracts, and your design rationale in a single session. It does not lose track of page 3 when writing page 15. Every section references the same grounded context, producing documents that read as a coherent whole rather than a series of disconnected paragraphs.
For the most demanding professional writing tasks, Opus is the benchmark. Heritage impact assessments that must balance conservation policy with design intent. Game design documents that maintain internal consistency across 50+ pages. Multi-stakeholder proposals where every claim must be traceable to evidence. When the document matters, Opus is the model to use.
Opus excels when a document must reference multiple source materials simultaneously. Feed it policy documents, site analysis reports, and design briefs, and it weaves them into a coherent narrative. Claims are grounded in the context you provide, not fabricated from training data.
Long documents are where most AI models fall apart. They contradict themselves, forget constraints, or drift from the brief. Opus maintains consistency across the full document length because it holds the entire context in working memory. Page 20 is as on-brief as page 1.
Start with Sonnet for most writing tasks. The quality gap between the two models is smaller than the gap between a good prompt and a bad one. Upgrade to Opus when you need massive cross-referencing, when the document exceeds 10 pages, or when internal consistency is critical. Opus costs more because it thinks harder.
A small indie studio building creative tools the way they should be built. No VC theatre, no funnel games, no faceless support.
Professional document drafting. Plans from £10/month.
Also available on Stensyl
Claude Sonnet
Natural prose. Fast drafting. Best value for most writing.
GPT-5.4
83% professional-level accuracy. Strong structure.
Gemini Pro
Native multimodal. Feed it images and data together.
GPT-5.4 mini
Near-flagship quality. Lowest cost per document.
Gemini Flash
Fastest writing model. 163 tokens per second.
Claude Sonnet
Natural prose. Fast drafting. Best value for most writing.
GPT-5.4
83% professional-level accuracy. Strong structure.
Gemini Pro
Native multimodal. Feed it images and data together.
GPT-5.4 mini
Near-flagship quality. Lowest cost per document.
Gemini Flash
Fastest writing model. 163 tokens per second.