Project Pricing

Project pricing sets a fixed fee for a defined set of design deliverables so the studio keeps the margin gains from AI tools instead of racing to the bottom on hours saved. You sell the outcome not the time. A client buys a complete brand system with prompt pack, token library shipped as code, and integration specs for their dev team. The price stays locked even if AI cuts your effort from 200 hours to 60. This model became the sweet spot for most AI augmented studios in 2026 because it prices the bigger deliverable clients actually need while protecting the efficiency no client needs to know about. 2026 US project bands run 8K to 25K for logo only work, 25K to 60K for full identity, 50K to 150K for brand systems, and 150K to 400K when you include the prompt library, Claude skills, and production ready tokens. EU numbers track 25 percent lower. LATAM and SEA sit at roughly half. The AI premium of 30 to 50 percent over 2024 comes from the expanded scope not from faster hours. Your proposal must list every asset, every format, every round of revisions, and every acceptance criterion or the model collapses under scope creep.

Project pricing is not hourly billing with training wheels. It is not the trap where you use Claude to deliver last years scope in one quarter the time then drop your price to match the new reality. It is not accepting hand wavy scopes like refresh our brand or make it pop. Those lead to endless feedback loops that destroy the margin AI promised. It is not keeping your prompt library and skills private while charging for outputs that clients cannot reproduce. Studios that fall into these patterns train their market to see AI as a reason to pay less instead of a reason to get more. The entire model only works when AI expands what you deliver beyond the 2024 ceiling of logos and PDFs. Anything less and you are just a faster version of your old self at lower effective rates.

Concrete proof lives in the 2026 engagement where Studio Lune took on the identity overhaul for the AI coding tool Cursor. The project kicked off with a 12 page scope document that listed 47 deliverables including full visual system, voice guidelines, 120 prompt pack for on brand asset generation in Midjourney and Claude, design tokens in TypeScript CSS and Tailwind configs, and an MCP server setup that let Cursor developers pull branded components directly. The fixed price landed at 215K with two revision rounds included and extras at 6K each. Lune used their internal skill library to accelerate motif exploration and consistency checks. Total time invested hit 82 hours versus the 220 hours the same scope would have required in 2024. Margin came in at 79 percent. Cursor walked away with systems their own team could extend which led to a 9K per month retainer for quarterly prompt and skill maintenance. The client never once asked about hours because the proposal focused on the expanded deliverables they could put to work immediately.

A smaller scale example shows how freelancers can use the same logic. Designer Alex Chen scoped a website UI and token system for the note taking app Reflect in late 2025. Old rate would have been 18K for basic screens and Figma files. With AI augmentation Chen quoted 42K for an expanded package that included responsive variants, motion specs, a Claude skill for design QA, and tokens that shipped as complete Tailwind config plus React primitives. The scope doc ran eight pages with 24 line items and clear handoff to GitHub. Work took 41 hours. Pre AI that same expanded scope would have taken 135 hours. Chen captured full fee and full margin while Reflect gained tools that cut their future design debt. Two months later Reflect referred a peer startup that signed at the same rate without negotiation.

Pull out project pricing when the work has clear start and end points, when you can write every deliverable on paper before ink hits contract, and when AI lets you ship systems that compound for the client. Brand systems with prompt packs, Figma MCP integrations that connect to Cursor or Claude Code, full design token libraries across five platforms, or one time Claude skill bundles fit this model perfectly. These projects let you command the higher bands because clients see the leverage they gain post handoff. Switch to retainers when the client needs ongoing support or when product direction changes monthly. Retainers run 8K to 20K per month for seed startups with defined surface area and response times. Avoid project pricing during early discovery phases where the deliverable is insight not artifacts. Bill discovery at top hourly rates of 200 to 250 in the US or roll it into the project as non optional phase one. Walk away when clients refuse to lock scope or when they fixate on hours saved instead of outcomes gained. Those clients will treat your AI efficiency as their permanent discount and your studio will never escape the race to the bottom.

Price the actual expanded deliverable AI lets you ship or watch your pricing power evaporate in the new market.