Support tiers, disclosed rates and a hidden price: packaging moved more than prices
Vapi rebuilt its plans around Success Packages, Shopify printed its Plus processing rates, Postman pulled Enterprise's $49 price, and Together raised Qwen3.7-Max 25%. 223 changes across 213 companies, 1.1 times the eight-week average.
This week in pricing · Sep 14 – Sep 20
2026W39223
+10% vs 8-wk avgpricing changes across 4,388 tracked pricing pages · 213 companies







The moves
Vapi replaced its two-tier Build and Scale pricing with a four-tier Success Package model: No Success Package at $0, Core at $29 a month, Pro at 10% of the hosting fee with a $999 monthly minimum, and Premier at a custom price, plus a new $20 per organization per month add-on [11]. The docs describe a Success Package as adding support, service commitments and higher included limits [1], and say that without one you use Usage only [2]. Zero Data Retention, previously a $1,000 a month add-on, is now included with Core and Pro; the entry-level concurrent call limit fell from 10 to 4, and regaining 10 means paying for Core [11].
Shopify put numbers where Plus used to say most competitive rates: online standard 2.25% plus 30¢, premium 2.95% plus 30¢, in-person card 2.3% plus 10¢, manual 3.4% plus 10¢, MB WAY and Bizum 3.25% plus 30¢, USDC 2.25% plus 30¢ [12]. The page keeps the third-party gateway rule, your processor's fees plus 0.20% per transaction to Shopify for security and compliance costs [3]. Buyer guides had treated Plus processing as a negotiated term discussed during the sales process [4]; the page now gives the list rate to negotiate from.
GitLab rewrote its Flex commitment FAQ: the question about volume discounts on overage was replaced by a reserved versus on-demand usage answer, two self-service questions were added for viewing the Flex balance and setting on-demand usage caps, and commitment allocation was narrowed from organizations or instances to Self-Managed instances only [13]. GitLab's own post frames adjustable spend caps as a way to keep teams from slowing down to renegotiate whose work gets priority [5]; the Flex dashboard is where the reservation and spend controls live [6].
Postman removed the Enterprise plan's public price of $49 per user per month billed annually and replaced it with Contact sales for pricing, and dropped the Get Started and Buy with AWS purchase options for Enterprise [14]. Third-party pricing pages still quote the old figure: one guide describes Enterprise at $49 per user monthly with 800 pooled AI credits per user [7], another lists Enterprise at $49 per user per month alongside Solo at $9 and Team at $19 [8].
Together AI raised Qwen3.7-Max serverless inference 25%: input from $2.00 to $2.50 per 1M tokens and output from $6.00 to $7.50 per 1M tokens, with cached input unchanged at $0.25 [15]. A catalog-maintenance thread had already flagged the previous step, from $1.25 and $3.75 to $2.00 and $6.00, as a stale entry [9]; routers now list the model at $2.50 in and $7.50 out per million tokens [10].
What they said
The vendor framing is about structure, not price. Vapi presents a Success Package as support, service commitments and higher included limits layered on usage [1]. GitLab sells its caps as budget guardrails so no team needs to slow down to renegotiate whose work gets priority [5]. Shopify states the one Plus processing rule it always published, the 0.20% third-party gateway fee, next to the newly printed card rates [3]. Postman says nothing: the price simply became Contact sales [14].
What the market said
The web is behind the pages this week. Postman buyer guides still carry $49 per user per month for Enterprise [7][8], a number the page no longer shows. Shopify Plus explainers still describe processing rates as terms discussed during the sales process [4], while the page now prints them. For Together AI, a catalog bot caught the previous Qwen3.7-Max increase and noted that Together's pricing page and docs disagreed on the amount [9]; routers have since settled on $2.50 and $7.50 [10]. For Vapi the docs are the only public source that explains the new tiers [1][2].
The pattern
Pricing pages logged 223 changes across 213 companies in the week of Sep 14 → Sep 21, 1.1 times the eight-week average of 194.9 [16]. Packaging led again: 92 packaging changes against 63 pure pricing changes and 68 product changes, with 80 expansions and 38 contractions [16].
Price increases were the one hot signal: 23 against an eight-week mean of 15.6, a 1.5 ratio, while cuts held at 7 [17]. Hidden-pricing events ran at 10 against 5 reveals [17], and Postman was the most visible of them [14]. Promotions ran at 21, a 1.2 ratio to the average of 17.8 [18].
The week's structural moves all shifted what is sold rather than what it costs: Vapi sells support tiers, GitLab sells spend control, Shopify sells a published rate card. Only Together AI moved a list price, and it moved it up [15].
What to watch
Whether Vapi customers on the old Build tier accept a 4-call concurrency limit or pay $29 for Core [11].
Whether Postman returns a public Enterprise price or keeps Contact sales [14].
Whether Together AI moves other Qwen models the same way after two Qwen3.7-Max steps in a month [9].
Whether Shopify publishes Plus processing rates outside the US pricing page [3].
Whether price increases stay above the eight-week mean next week after a 1.5 ratio [17].
Written with 85% of the week’s diffs published; late-published changes appear in the feed.
Sources
- [1]Vapi docs — Pricing and Success Packagesverified“A Success Package adds support, service commitments, and higher included limits.”
- [2]Vapi docs — Pricing and Success Packages (usage only)verified“Without an optional Success Package, you use Usage only.”
- [3]Shopify Plus — pricingverified“You pay your processor fees, plus 0.20% per transaction to Shopify for security and compliance costs.”
- [4]Swell — Shopify Plus Pricing in 2026verified“Unlike standard Shopify plans with published, non-negotiable rates, the Plus tier operates more like enterprise software where terms are discussed during the sales process.”
- [5]GitLab blog — See who spent your AI credits and set fair caps per teamverified“As consumption changes, you can adjust spend caps, so no teams need to slow down to renegotiate whose work gets priority.”
- [6]GitLab docs — GitLab Flex Usage dashboardverified“The Flex dashboard is where you track your GitLab Flex commitment and usage, set your monthly reservation, and control your spend.”
- [7]Apidog — Postman Pricing 2026: What They Changedverified“The Enterprise plan at $49 per user monthly (billed annually) adds 800 AI credits per user (pooled), API Catalog, unlimited Partner and internal workspaces, Private API Network, Advanced RBAC and organization controls”
- [8]CostBench — Postman Pricing 2026verified“Plans: Free (free), Solo at $9/user/month, Team at $19/user/month, and Enterprise at $49/user/month.”
- [9]Braintrust proxy issue #1245 — Together: update stale Qwen/Qwen3.7-Max pricingverified“The current catalog entry shows $1.25 input / $3.75 output per million tokens, but Together AI's pricing page now shows $2.00 input / $6.00 output.”
- [10]Requesty — qwen3.7-max API pricingverified“qwen3.7-max is priced at $2.50 per million input tokens and $7.50 per million output tokens when accessed via Requesty.”
- [11]Pulse capture: Vapi scraps Build/Scale for four-tier Success Packages from $0 to custom“Vapi scraps Build/Scale for four-tier Success Packages from $0 to custom”
- [12]Pulse capture: Shopify: Plus payment processing rates now publicly disclosed“Shopify: Plus payment processing rates now publicly disclosed”
- [13]Pulse capture: GitLab: Flex billing FAQ overhauled — new usage-cap self-service, terms narrowed“GitLab: Flex billing FAQ overhauled — new usage-cap self-service, terms narrowed”
- [14]Pulse capture: Postman hides $49/user Enterprise price behind Contact Sales, drops AWS buy“Postman hides $49/user Enterprise price behind Contact Sales, drops AWS buy”
- [15]Pulse capture: Together AI: Qwen3.7-Max inference price increased 25% ($2.00→$2.50)“Together AI: Qwen3.7-Max inference price increased 25% ($2.00→$2.50)”
- [16]Pulse weekly stats, 2026W39Pulse Index data“223 changes across 213 companies; 92 packaging, 63 pricing, 68 product; 80 expansions, 38 contractions; eight-week mean 194.9, ratio 1.1”
- [17]Pulse weekly stats, 2026W39Pulse Index data“23 price increases (ratio 1.5 to the 15.6 mean), 7 price cuts, 10 hid pricing, 5 revealed pricing”
- [18]Pulse weekly stats, 2026W39Pulse Index data“21 promo events, ratio 1.2 to the 17.8 mean”
Questions about this week?
The Pulse Agent has every change from this issue loaded — the story, the numbers and each company's diff — and answers from verified data.
Ask the agentBy the numbers
- pricing changes tracked· about normal
- 223
- price increases vs cuts
- 23 ↑ / 7 ↓
- expanded vs contracted plans
- 80 / 38
- packaging moves
- 92
- discount & promo moves· above normal (+20%)
- 21
In this issue
Previous issues
All weekly issues →- Week 41, 2026·Sep 28 – Oct 4·279 changesIntercom's $19 seat is $29 again, five days after the cut
- Week 40, 2026·Sep 21 – Sep 27·276 changesPackaging rewrites outpace price moves: credits, bundles and metered add-ons
- Week 38, 2026·Sep 7 – Sep 13·196 changesCredits replace rate limits, flat rates replace meters: a normal week with structural moves
- Week 37, 2026·Aug 31 – Sep 6·204 changesMetrics and packages rewritten: steps, hosts, hours and regions outnumber price hikes
- Week 36, 2026·Aug 24 – Aug 30·149 changesA quiet week for list prices, a loud one for metering: tokens, seats, events and parsing
- Week 35, 2026·Aug 17 – Aug 23·209 changesAgent pricing takes shape: usage models, factory tiers and platform fees replace flat plans