Braintrust pricing teardown
https://www.braintrust.dev/pricingBraintrust's pricing page is a well-structured usage-based model with a generous free tier, a clear mid-tier at $249/mo, and an enterprise contact tier — covering the full spectrum from solo developers to large teams. The page includes a live usage calculator and an FAQ, which are strong trust and conversion tools, but the absence of an annual discount option and a non-featured/non-highlighted recommended plan leave meaningful conversion optimization on the table.
Tier structure
Starter
$0/mo
Pro
$249/mo
Enterprise
Custom
Value metric
usage-based flat tier with overage charges (scores, processed data, Topics tokens)
How limits scale
Escalation logic
Tiers escalate from a free entry tier with tight usage limits to a $249/mo Pro tier with 5x more included usage and team/compliance features, then to a fully custom Enterprise tier for high-volume or privacy-sensitive deployments. Upgrade triggers are primarily usage volume (scores, data), compliance needs (RBAC, DPA, SSO), and support SLAs.
Psychological anchors
- Free tier as entry anchor — Starter at $0 with no credit card required lowers the barrier to entry and is explicitly called out twice ('No credit card required') to reduce signup friction.
- Strikethrough/discount signal on Pro — The page text shows '$100' crossed out next to '$249' on the Pro tier, suggesting a prior or promotional price — though the nature of the discount is unclear without visual context.
- Startup program as soft discount — '6–12 months free for qualifying startups' is displayed on both the Pro card and the CTA section, acting as a conversion hook for early-stage teams who might otherwise stay on Starter.
- Enterprise as high-anchor — Custom pricing with no public number anchors Pro at $249 as the 'reasonable' option by contrast, a classic decoy/anchor effect.
- Interactive usage calculator — A live calculator with preset use-case scenarios (Support agent, RAG assistant, etc.) lets users self-qualify and see estimated costs, reducing sticker shock and building confidence before upgrade.
- Unlimited non-scarce resources — Users, projects, datasets, playgrounds, and experiments are all Unlimited across all tiers, framing the value metric purely around usage volume rather than seat count — removes a common growth penalty.
What this page is optimizing for
The page is primarily optimized for self-serve conversion at the Starter and Pro tiers, with the free tier, no-credit-card CTA, usage calculator, and startup discount all designed to move developers from zero to paid quickly. Enterprise lead-gen is secondary, handled via a simple 'Contact us' CTA with feature differentiation (SSO, BAA, custom retention) doing the qualifying work.
Red flags
- No annual billing option is visible anywhere on the page, forfeiting the standard 17–20% discount lever that drives annual commitment and reduces churn.
- No plan is visually highlighted or badged as 'Most Popular' or 'Recommended' — without a screenshot confirming a visual highlight, the decoy effect between Starter and Pro is weakened.
- The strikethrough price on Pro ($100 crossed out, $249 shown) is confusing — it implies Pro costs more than a prior price, not less, which is the opposite of a typical discount signal and may cause trust friction.
- The $249 Pro platform fee is not offset by the included $249 Topics credits in a clearly communicated way, making the net cost opaque for users who primarily use Topics.
- Human review scores capped at '1 per project' on Starter is a meaningful limit that could frustrate teams doing annotation workflows, but it's buried in a feature comparison table rather than surfaced as a key upgrade trigger.
- FAQ section exists but answers are not visible in the page text, suggesting they may be collapsed accordions — if answers are thin or generic, this is a missed objection-handling opportunity.
- No social proof metrics (e.g., number of teams, evals run, traces processed) are present beyond 'Trusted by the best AI teams,' which is generic without logos or numbers visible in the text.
Best-practices scorecard
What works · 3
- Value metric matches usage — Charging on scores, processed data, and token consumption directly maps to how AI evaluation teams actually scale their usage.
- Tier differences are scannable — A detailed feature comparison table with clear row-by-row differences across all three tiers makes differentiation easy to scan.
- Clear CTAs per tier — Each tier has a distinct CTA ('Sign up', 'Get Pro', 'Contact us') that matches the appropriate conversion path for that tier.
Half measures · 3
- Visible prices — Starter ($0) and Pro ($249) are clearly priced, but Enterprise is 'Custom' with no range or starting price, and the Pro strikethrough creates ambiguity.
- FAQ or objection handling — An FAQ section is present with relevant questions (plan selection, billing, discounts), but answers appear to be hidden behind accordions and are not readable in the page text.
- Trust signals present — 'Trusted by the best AI teams' is present but lacks logos, customer counts, or testimonials in the visible text; SOC 2 Type II compliance is mentioned in the feature table which helps.
What's missing · 2
- Clear recommended tier — No 'Most Popular' or highlighted plan is detectable from the page text; Pro is not explicitly positioned as the default choice.
- Annual discount offered — No annual billing toggle or discount is mentioned anywhere on the page.