Skip to content
Back
Pricing briefing
CCognition
Cognition
PRICING METRIC

Cognition flattened its Teams plan pricing

By , Co-founder, PricingSaaS

Published

Cognition replaced Teams' per-seat fee with one flat monthly price.

Cognition’s pricing page · Change detected

What changed

Cognition's Teams plan dropped its $40-per-month, per-full-developer-seat fee and now bills a flat $80 per month, according to Cognition's pricing page as of Nov 18, 2026 — the day Pulse detected the change. The plan comparison table's "per full user" footnote, which spelled out the old per-seat math, was removed along with the fee.

Cognition also added "Devin Bots" as a new capability across the Max, Teams, and Enterprise plans, per the same pricing page update: it now appears on each plan's card and as a new row in the comparison table, listed under Devin Cloud.

Why it matters

Flattening the Teams plan removes a variable that made per-seat budgeting unpredictable for growing engineering teams: instead of paying more every time they added a full developer seat, Teams customers now pay one fixed monthly price no matter how many seats they add. That favors teams that are scaling up seats and is a worse deal for teams that were already small, since they lose whatever savings came from paying for only the seats they used.

It also lands seven days after Cognition halved the Team plan's user cap to 100 and added ChatGPT plan support, per Pulse's tracking of that change on Nov 11, 2026 — pulling in different directions (fewer seats allowed, but no extra charge per seat) and suggesting Cognition is still actively shaping how Teams customers pay rather than settling into a model. Pairing the fee removal with a new Devin Bots capability also signals where Cognition wants the value story to shift: from how many human seats a team buys to what its AI agents do.

Watch whether the flat fee survives as teams grow past a handful of seats, or whether Cognition reintroduces a per-seat charge at a new tier once usage climbs — given how fast its pricing has moved this year, another adjustment is plausible.

In context

Cognition's own history with seat pricing is thin: Pulse has recorded just one earlier move in this pattern, capping the Teams plan at 200 members on Sep 24, 2026 [[cognition/2026D267]], so a flat-fee swap doesn't read as a settled model yet. The faster signal is cadence — Pulse has tracked five pricing or packaging changes to Cognition in the last 90 days, averaging about one every 21 days since it began tracking the company on Jul 6, 2026, which points to a team still iterating rather than optimizing one lever. Seat pricing itself is crowded elsewhere: Pulse tracked 175 seat-pricing changes across 126 companies in the last 180 days, 5.3% of the 3,298 changes it tracked in that window, with Invideo and Intercom among the most active in it. Most of those peers add seat tiers or reprice per-seat rates rather than removing the charge outright, making Cognition's flattening — paired with the new Devin Bots capability — read less like a discount and more like a bet that usage-based value matters more to Teams buyers than headcount.

Common questions

How do we size this for our own Teams plan?

There's nothing to multiply any more: Cognition's Teams plan now charges one flat rate per month regardless of how many full developer seats you add, so sizing it is just comparing that flat rate to what your team was paying for its current seat count under the old formula.

Is this part of a broader pattern, or a one-off?

Treat it as one data point in an unusually fast stretch of change — Cognition has touched its pricing and packaging repeatedly over the past few months, so don't assume the flat fee is final before locking in a long-term seat-based deal.

Should we consider flattening our own per-seat fees?

Only if usage, not seat count, is the better predictor of value for your product — Cognition paired the fee removal with a new Devin Bots capability, suggesting it wants usage-based features rather than seats to carry its pricing story going forward.

Sources