Lessons From Lovable: Pricing For AI - Elena Verna
Elena Verna had a webinar with Metronome yesterday and shared her learnings at Lovable.
Lessons from Lovable: Pricing for AI.
Elena was as brilliant as she always is, and the notes below are gold.
Lovable’s Current Pricing Model
Four tiers: Freemium → Pro ($25) → Business ($40) → Enterprise
Credit-based system with monthly allocations:
Freemium: 5 credits daily, 30 total per month
Pro: 100 credits / month (upgrade options: 200 or 500)
Business / Enterprise: Volume pricing with additional features
Key differentiators by tier:
Custom domain publishing (Pro+)
Remove Lovable branding (Pro+)
Roles & permissions (Business+)
SSO & data opt-out (Business+)
Credit system: Philosophy and Challenges
The main driver for customers to upgrade their plans on Lovable is the need for more credits rather than additional features.
Customer pain points with credits:
Unpredictable cost per action
Hard to compare pricing across competitors
Mismatch between expected vs actual credit consumption
Credit rollover policy:
Retention tactic: if a customer churns, they lose rollover balance
One-month forward only, expires if unused
Exploring alternatives:
Top-up purchases (testing in January)
Potential “Lovable wallet” with direct dollar spend
May move away from credits long-term if LLM costs decrease
Strategic Pricing Decisions
Removed per-user pricing to unlock collaboration
“We want to unlock as many input metrics as possible and monetize on output metric.”
“Unlimited users, and you buy just credits on the workspace.”
Migrated team plan users from $40 to $25 / month; Took ~$4M revenue hit but improved retention and collaboration ~10x
Freemium as marketing expense, not cost center
Gives away credits for hackathons, education, nonprofits
Lovable badge on free apps drives viral growth
Focus on engagement over revenue optimization
Goal: maximize number of paid users, not ARPU
Pricing changes and experiments aim to drive engagement more than revenue
Constantly exploring how to make the product cheaper
Daily 5 credits on top of monthly allocation
Encourages users to return and make incremental progress, building a habit loop
“Ingenious growth hack” for maintaining engagement
Elena’s view: should probably be in every product
Enterprise & market expansion
Enterprise plan launched 4 months ago due to strong demand
Requests for:
Company-wide solutions
Added security features
Many “enterprise” features live in Business plan
Supports SMB / mid-market (sub-1,000 employees), and they can avoid agreement for enterprise
“You can see SSO is our self-serve feature, not enterprise feature.”
Self-serve → enterprise motion
Enterprise clients start self-serve
Once value is proven, upsell to enterprise agreements
Regional pricing experiments
$5 entry point for India, Brazil, Indonesia (January test)
Credit allocation sophistication
Exploring individual vs pooled credits for teams
Experimentation & future outlook
Dynamic pricing as a norm
Frequent changes help normalize pricing evolution with customers
Test on new users first, then migrate existing
New users are the future
Existing users have a fixed perception of pricing → risk of negative feedback
Willingness to take short-term revenue hits
For more user-friendly changes (price reductions, feature unlocks)
Next 12–24 months in AI pricing
Market moving away from simple pass-through LLM pricing
Pass-through (LLM cost + ~20% margin) = defensive, immature market
Credits as an interim solution
Likely temporary until LLM costs decrease significantly
AI startups forced to optimize pricing early
“Usually it takes companies about five years before they start doing a lot of price and packaging work; almost all AI startups are forced to start in their first year.”