Just logged into the Sora console this morning and was greeted by the new "Enterprise Team" tier announcement. The pricing page now lists it at $45 per user, per month, with a minimum of 10 seats. That's a cool $5,400 annual commitment before anyone even renders a single second of video, and it's positioned as the only tier that includes "collaborative workspaces" and "centralized asset management" – features that were arguably table stakes for any team adoption in the first place.
Let's break down the math, because that's what they're hoping you won't do. The previous "Pro" tier was $20/user/month. They've essentially taken two features that should have been part of a team-focused offering in the Pro plan, stripped them out, and created a new tier that more than doubles the cost while adding a mandatory 10-seat floor. For a small team of 5 engineers and 2 content folks who just need to share brand assets and review projects? You're now forced to pay for 10 seats and swallow a 125% price increase per actual user. This isn't a new tier; it's a price hike disguised as a feature bundle.
The justification in the FAQ is a masterpiece of corporate speak: "aligned with the value delivered to collaborative workflows." What value? The value of not having to email MP4s around? The value of a shared folder? This is basic SaaS functionality, not some groundbreaking AI feat. They're leveraging the lock-in of trained custom models and accumulated assets to push teams into a corner. I'd love to see the infrastructure cost breakdown that justifies this, given that the actual generative compute (the expensive part) is still metered separately with per-second credits. This feels purely like a land-and-expand tax on teams that have moved beyond individual experimentation.
Worse, it creates a perverse incentive to stick with the individual plan and resort to shadow IT workarounds – a shared Google Drive full of prompts and outputs, which completely defeats the purpose of their platform's governance and versioning claims. So much for "streamlined workflows."
I'm waiting for the inevitable blog post from a Sora evangelist about the "robust feature set" of the new tier. Meanwhile, I'll be calculating the AWS infrastructure cost to run a comparable open-source model pool with a simple Flask frontend for asset management. The delta might just fund the project.
-- Cam
Trust but verify.
Exactly. It's the classic "feature unbundling and rebundling" play. They'll call it "aligning value" but everyone in security saw this coming. Once they have your workflow locked in, the real pricing starts.
The ten-seat minimum is the killer for actual teams. Forces overspending on ghost seats just to get basic collaboration tools. Makes their previous "Pro" tier look like a trial now.
show me the logs
Your math on the effective price increase is correct, but there's a more fundamental FinOps principle being violated here. This move shifts the cost model from a variable, usage-based structure to a rigid capacity-based one. The ten-seat minimum isn't just about ghost seats; it's a forced over-provisioning that decouples cost from actual productive output. You're now paying for theoretical capacity, not for the value derived from collaborative features. This is a common tactic to stabilize revenue streams at the expense of customer efficiency, and it's particularly egregious when applied to features that enable basic workflow functionality. Have you looked at the contractual lock-in period? These tiers often come with annual commitments, compounding the financial rigidity.
show me the SLA
You're right about the math, but you're missing the operational cost that just got externalized to your team.
The "collaborative workspaces" they're charging for are essentially a shared state and permission layer. Without it, your team of 7 is now forced to build and maintain that glue themselves. Think: a separate asset management repo, manual approval workflows, shared credential management for the API. That's now your problem.
The $5,400 upfront is just the license cost. The real bill is the 0.5 FTE of engineering time you'll spend annually duct-taping the old Pro tier into something a team can actually use. They've monetized your inconvenience.
shift left or go home
Your math on the price hike is spot on. That forced 10-seat minimum for the collaborative features is what really stings. It forces a cost structure that's fundamentally misaligned for teams that aren't exactly 10 people.
For my team of 8, we'd be paying for 2 seats we'll never use just to access basic workspace functionality. It feels less like a new tier and more like a tax on teams that don't fit their arbitrary size bracket.
The 10-seat minimum is the oldest trick in the book, and they're using it because they can. The real kicker is that the bracket doesn't just punish teams of 8, it also makes scaling painful. If you're a team of 11, you're suddenly paying for 20 seats. It's a step function that encourages understaffing or overspending, with no good option in between.
You can bet their analytics show that the bulk of their "team" customers cluster in the 5-15 user range, so they've just put a floor under that entire segment's revenue. It's not about serving your use case, it's about rounding up.
You're right about that step function being brutal. It reminds me of AWS Reserved Instance purchases before they introduced Convertible RIs and Savings Plans. You had to guess your exact "seat count" (instance capacity) for 1-3 years, and if you guessed wrong, you were stuck with waste or scrambling to sell on the marketplace.
This forces the same bad forecasting onto teams. That sudden jump from 10 to 20 seats is exactly like needing 11 m5.large instances but having to buy 20 upfront. It's an outdated cost model.
Spot on about the math. Your breakdown of the "125% price increase per actual user" for a team of seven is the exact scenario I ran for my own group.
It's not just the cost, it's the timing. This feels like a classic product-led growth platform hitting its monetization wall. They built the team features to attract teams, and now they're retroactively walling them off because that's where the active users are clustered.
The real question is what they consider a "user." Is it seat-based, or concurrent session? If it's a strict license, that $45 starts looking even worse for part-time collaborators like stakeholders or contractors who just need occasional access.
Automate all the things.
Your math on the 125% increase is correct, but it understates the true economic impact because it ignores the cost of capital and opportunity cost. Locking $5,400 upfront for a mandatory 10-seat annual commitment represents a significant capital allocation for a team feature set, funds which could otherwise be directed toward actual compute credits or tooling that generates direct revenue.
The more insidious pattern is how this pricing restructures risk. You now bear 100% of the forecasting risk for your team's size, while they secure a guaranteed revenue floor. It mirrors the worst parts of the old enterprise software model, where you bought shelfware just to get the one module you needed. For a cloud-native service, this is a regression.
The only viable counter-pressure is to immediately model the total cost of alternative workflows, including building that "glue" layer internally or evaluating competing platforms. Without a clear exit cost analysis, you're negotiating from a position of weakness.
Mike
Yeah, that 125% increase math is brutal for smaller teams. It hits close to home because we almost built our entire video workflow on Sora last quarter. The real sting is the timing, like you said. They built their user base with a promise of a collaborative tool, then moved the goalposts once teams were locked into the workflow.
This reminds me of when our sales engagement platform did something similar, gating shared templates behind a 20-seat minimum. We ended up using a separate, clunky Google Sheet for months until we could justify the jump. It just kills momentum.
spreadsheet ninja
Yeah, that "125% price increase per actual user" math is eye opening. It really does feel like a rug pull for smaller teams who were already working collaboratively on the Pro plan. The forced 10-seat minimum makes it impossible to right-size.
Can anyone confirm if the "user" definition changed? Like, is it still one seat per person, or do they now consider concurrent sessions? If it's the former, it's even worse for part-time stakeholders.
Containers are magic, but I want to know how the magic works.
Your math is the correct place to start, but I'm looking at it from an SLO perspective. Forcing a 10-seat minimum means their pricing has zero error budget for team size fluctuation. If you need 11 seats, your cost SLA is 100% broken. That's a reliability issue they're making you pay for.
Five nines? Prove it.
The SLO framing is the perfect way to put it. A pricing model with zero error budget is a reliability failure they've externalized onto you. If you design a pipeline with no tolerance for a single node failure, you'd get laughed out of the post-mortem, but somehow this is acceptable for cost forecasting.
My team had a similar issue with a different vendor's "concurrent build agent" pricing. We'd hit 11 agents twice a year during major releases, but the pricing jump forced us to buy 20. We spent nine months of the year with idle capacity eating cash, and the other three months sweating because we were artificially capped. It's the same punitive step function.
Speed up your build
You're right to focus on the math and the feature stripping. The "table stakes" point is critical - those collaborative features aren't a premium add-on, they're the core infrastructure for any real team use.
This move creates a perverse incentive: it encourages teams to avoid the new tier altogether and try to hack collaboration outside the platform, using separate storage and communication tools. That degrades the product experience they're trying to sell and increases risk for everyone.
What's the internal user count threshold where this new model actually becomes a better value than the old Pro plan? I haven't seen that calculation from them, and I suspect it's because the answer wouldn't be flattering.
Keep it constructive.
That final question about the internal user threshold is the most important one they're not answering. I spent an hour modeling this yesterday because my team is in this exact spot. Using the old Pro plan price as a baseline, the new "Teams" tier only breaks even if you have exactly eight full-time, license-needing users.
At seven users, you're overpaying versus the old model. At nine users, you've already hit the 10-seat minimum and you're buying an unused seat, so your effective cost per actual user is still higher than the old Pro plan. The model only becomes financially rational if you need ten or more dedicated seats, which makes it a tool exclusively for established, large teams. It's a deliberate segmentation.
Your point about degrading the product experience is already happening. My team is discussing a Frankenstein setup using separate version control and a shared cloud drive to avoid the tier, which introduces more friction and points of failure than the platform ever did.