# Per-Seat Pricing > Source: https://ibl.ai/resources/glossary/per-seat-pricing > Last updated: 2026-08-19 **Definition:** Per-seat pricing charges a fixed fee for every licensed user per period, regardless of how much any individual actually uses the product — so cost scales with headcount rather than with consumption. **On ibl.ai you own all the code and the data, run it model-agnostic across any LLM, and pay with no per-seat pricing — so you can deploy anywhere, from your own cloud to a fully air-gapped network.** ## What is Per-Seat Pricing? The model assumes uniform consumption. AI consumption is not uniform; it follows a power law, with a minority of users generating the large majority of requests. Under seat licensing, the engineer running 400 requests a day and the colleague who logged in once during onboarding cost exactly the same. The advertised number is also frequently not the real one. AI features are often sold as an **add-on** requiring a qualifying base licence underneath, so the sticker understates the committed spend — Microsoft 365 Copilot is $30 per user per month as an add-on and roughly $69 (E3) or $90 (E5) all-in once the base plan is counted. The structural consequence is that efficiency does not pay. Halve your token consumption through better routing or prompting and the invoice is identical; the saving accrues to the vendor. ## Why It Matters Per-seat pricing determines whether an AI deployment can scale past a pilot. At small headcount the difference against usage-based pricing is immaterial; above roughly a thousand users the two curves separate by an order of magnitude for identical work. ## Key Characteristics ### Cost Tracks Headcount, Not Use The bill rises when you hire and falls when you make redundancies. Neither movement has any relationship to how much AI work the organization actually performed. ### The Sticker Is Often an Add-On Price AI seats frequently require a qualifying base licence underneath. Counting it can double or triple the real per-seat figure before any agent consumption is added. ### Efficiency Gains Accrue to the Vendor Reduce consumption by half and the invoice does not move. Under usage-based pricing the same optimization is a permanent reduction in your own cost. ### Dormant Seats Are Paid in Full Because consumption follows a power law, a large share of licensed seats generate negligible usage while costing the same as the heaviest users. ### Promotional Rates Carry Expiry Dates Introductory seat pricing is common and time-boxed. The renewal figure, not the promotional one, is what belongs in a multi-year business case. ### Renewals Entangle With Suite Licences When the AI seat is an add-on to a productivity suite, the AI renewal is bundled with a licence renewal the organization was never going to walk away from. ## Examples - **Enterprise:** A 10,000-employee organization budgets an AI rollout from a $30 per-user add-on price without counting the required base licences. — *The real all-in figure is $69–$90 per seat per month, or $8.3M–$10.8M a year against a $3.6M budget — a gap large enough to fail the business case.* - **Enterprise:** An organization improves prompt efficiency and routes routine work to a cheaper model, cutting token consumption substantially. — *Under per-seat licensing the invoice is unchanged. The engineering effort produced margin for the vendor and nothing for the buyer.* - **Public Sector Agency:** A public agency must license every employee for compliance reasons even though AI use concentrates in two departments. — *The majority of the spend buys dormant capacity, which is difficult to defend in an appropriations review that asks what the money bought.* ## Why does ibl.ai have no per-seat pricing? Because cost should track work, not headcount. ibl.ai is the agentic AI platform where you own all the code and the data, and it carries no per-seat pricing: you pay for usage against a budget cap you set, with credits pooled across every user, model and agent rather than allocated per person. The ten-thousand-and-first user costs nothing until they actually run something. Because it is model-agnostic you can route routine work to a cheaper model and keep the saving, and because you hold the source under a perpetual license every efficiency gain reduces your own bill permanently. You can deploy anywhere. 1.6M+ users across 400+ organizations run the platform this way, including NVIDIA, MIT, and Syracuse University. ## FAQ **Q: What does Microsoft 365 Copilot actually cost per seat?** The advertised $30 per user per month is an add-on price requiring a qualifying Microsoft 365 base licence. All-in it is approximately $69 per seat on E3 or $90 on E5 following the 1 July 2026 base-plan increase, before any Copilot Studio agent consumption. **Q: At what size does per-seat pricing stop making sense?** There is no universal threshold, but the curves separate as headcount grows and usage concentration increases. The practical test is to compare the seat bill against the token cost of the work actually performed — if the ratio is several multiples, the pricing shape is wrong for you. **Q: Is usage-based pricing riskier because the bill is unbounded?** Only if the platform has no ceiling. A well-designed usage model has a budget cap the organization sets, with auto-refill as an explicit opt-in rather than a silent default, so the maximum is a number you chose rather than one you discover. **Q: Do seat prices fall as inference costs fall?** Historically no. Inference cost per token has declined steadily while seat prices have been stable or risen. Under a flat per-seat fee that decline is vendor margin; under usage-based pricing on infrastructure you own, it is your saving. **Q: How do you compare per-seat and usage-based pricing honestly?** Model the same workload under both: headcount times the all-in seat rate on one side, measured token consumption times the model rate on the other. Include base licences in the seat figure and infrastructure in the usage figure, or the comparison flatters one side. ## How does ibl.ai approach Per-Seat Pricing? **ibl.ai is the agentic AI platform where you own all the code and the data. You self-host the entire stack inside your own perimeter, run it model-agnostic across any LLM and switch anytime, and pay by usage with no per-seat pricing — so you can deploy anywhere: your cloud, on-premise, GovCloud, or fully air-gapped.** - **You own all the code and the data.** Full source code under a perpetual license, running on your infrastructure. Not API access to someone else's platform — the stack itself is yours. - **Model-agnostic.** Run any LLM — Claude, GPT, Gemini, Llama, Command, or your own fine-tune — and switch providers without rewriting the platform. - **No per-seat pricing.** Usage-based billing against a budget cap you set. Cost tracks what your organization actually uses, not how many people you employ. - **Deploy anywhere.** Your cloud, your VPC, on-premise, GovCloud, or a fully air-gapped network with no outbound connectivity. 1.6M+ users across 400+ organizations run the platform this way, including NVIDIA, MIT, and Syracuse University. ibl.ai is family-owned and operated from New York, NY — a U.S.-headquartered, domestically-owned long-term partner, not a vendor that sells licenses and moves on.