Built by a food scientist, for food scientists
Brix is the software food R&D teams use instead of spreadsheets. Enter your ingredients and recipes, and Brix builds your nutrition facts, ingredient statements, BOMs, batch sheets, and costs automatically. It matches how you already work. Manage projects, keep lab notes, and compare formulas side by side, all in a clean intuitive workspace.

Meet Rigatoni, the AI built into Brix. Ask him anything about your formulas, costs, and specs and move faster.
Free 14-day guided pilot on your own data, after a short demo. Pricing is public, right here on this page.
Watch a recipe get built, costs cascade, and a label generate, start to finish.
For the formulator
When you update ingredients and recipes in Brix, everything connected to them updates too: nutrition facts panels, ingredient statements, costs, and your bill of materials. No spreadsheet formulas required.
The part that took an afternoon is now a few clicks.
For the R&D manager
No more emailing around for the current versions of formulas. Your whole team works in one system, on the same data, always the current version.
R&D knowledge stays in the system, not in someone's inbox.
For regulatory & QA
Nutrition Facts follow 21 CFR 101.9: rounding rules, %DV math, serving declarations, allergens. Nutrient values come from data you control: supplier-provided nutrition specs you enter, or the USDA reference database. Export as a vector PDF or an image, in all the FDA formats.
Every value traces back to its source, so your QA team can sign off with confidence.
AI Native
Most R&D tools bolted on an AI tab. Brix was built differently: AI inside every workflow, and your data open to any model or agent you bring.

Rigatoni, the Claude-powered assistant built into Brix, creates recipes, calculates nutrition, runs what-if cost scenarios, generates labels, and reverse-engineers competitor products.
Unlike standalone AI tools, Rigatoni works on your actual data: your ingredients, your suppliers, your recipes, your costs.
Export anything to CSV or PDF, anytime. No walled garden, no lock-in.
Use Rigatoni, or bring your own AI: Claude, ChatGPT, and in-house tools connect via API and MCP. We don't lock you into one model or one workflow.
Most formulation tools charge separately for the label module, the API, extra seats, and support. Brix is $250 per seat per month, billed annually: every feature, no add-ons, no separate modules to buy.
We're not cheaper because we're missing things. We're cheaper because we didn't build a 30-year-old pricing model.
Billed annually — $3,000 / seat / year
Book a demoFree 14-day pilot after your demo. No credit card.
Enterprise, custom pricing: custom integrations with your organization's existing systems. Contact us
Yes. The two-minute video above is real product, no form required. When you're ready, book a 30-minute demo and we'll set you up with a free 14-day pilot on your own ingredients and formulas, helping you get productive so you're evaluating on real data. Our pricing is public on this page, so the demo is about getting you set up, not prying a quote out of you.
ChatGPT doesn't know your ingredients, your supplier prices, or your recipes. Rigatoni, the AI built into Brix, works on your actual data. Ask "what's the cheapest way to hit 5g of protein in this formula" and it answers from your ingredient library and your costs, not the internet. And it does the work, not just the talking: builds the formula, runs the cost scenario, generates the label.
Yes. The calc engine implements FDA 21 CFR 101.9: rounding rules, %DV math, mineral thresholds, allergen declaration logic. Nutrient values come from supplier specs you enter or the USDA reference database. Export the panel as a vector PDF with embedded fonts or as an image, in all the FDA formats.
Yes. One retail SKU can have multiple plant-specific formulas with shared label settings. Update a co-man's formula without breaking the parent product.
Yes. Upload your ingredient master, recipes, and supplier specs as CSV, and Brix matches them automatically. Or drag in supplier spec sheets and let Rigatoni extract the data. Most teams get a working library running the same day.
Your data is US-hosted, encrypted at rest and in transit, and isolated from other companies' data. It is never used to train AI models. Export everything anytime via CSV, PDF, or API, and take it with you if you cancel.
Built by a food scientist who got tired of the spreadsheet stack.
$250/seat/month, billed annually. Full stack, no hidden tiers, no add-on modules.