Google Gemini enterprise pricing & licensing.
Gemini reaches the enterprise on two separate commercial planes: a per-seat Workspace add-on and consumption-metered Vertex AI. This note maps the list prices, the Workspace bundling lever, the Vertex token benchmarks, and the negotiation moves that move Google’s field teams — so you stop paying twice for AI you already own.
There are two Gemini pricing planes, not one: a $30/user/month Workspace add-on and consumption-metered Vertex AI. The single highest-value lever is folding the Workspace add-on into a multi-year Google Cloud commitment — that moves list $30 to $24–$27 and unlocks 15–25% blended discounts. Buyers who negotiate the add-on in isolation rarely beat 5–10%.
01 Key findings
Two pricing models, not one. Gemini for Workspace is per-seat; Gemini in Google Cloud (Vertex AI) is per-token. They are priced, contracted and billed separately, with no unified identity or billing plane between them.
Gemini is never bundled into a base Workspace tier. The $30/user/month add-on sits on top of Business Standard ($14) — a $44/user/month floor before you touch Google Cloud. At 500+ users that is $250,000+ a year.
The overlap trap is real. Enterprises already on Vertex AI frequently double-pay — layering a per-seat add-on over capability their applications already consume through the API. Google’s sales teams rarely surface it.
Discount only travels with commitment. The deepest cuts (15–25% off list) are reserved for Workspace + Cloud + security stacked into one multi-year deal. Single-product add-on negotiations top out at 5–10%.
Consumption is the unmanaged risk. Vertex defaults to permissive quotas; without caps, committed use discounts (CUDs) and budget alerts, token metering produces mid-quarter bill shock.
02 Workspace pricing
Gemini for Google Workspace (formerly Duet AI) is a standalone add-on available only to Business Standard, Business Plus and Enterprise subscribers — $30 per user/month a la carte, or $360 per user annually. It layers writing help in Docs, drafting in Gmail, intelligence in Sheets, design help in Slides and Meet summarisation onto the base seat.
| Component | Unit | List price | Gemini eligibility | Note |
|---|---|---|---|---|
| Gemini for Workspace add-on | per user / mo | $30 ($360/yr) | — | Formerly Duet AI; a la carte |
| Business Standard (prerequisite) | per user / mo | $14 | Eligible | Minimum tier — creates a $44 floor |
| Business Plus / Enterprise | per user / mo | Higher | Eligible | Add-on applies at same $30 |
| Business Basic | per user / mo | $6 | Not eligible | Forced upgrades can add ~50% to budget |
03 Vertex AI API pricing
On the Cloud side, Gemini is metered by input and output tokens through Vertex AI. Unlike the fixed per-seat add-on, consumption is unpredictable and scales with usage — a single heavy application can burn $5,000–$20,000 a month. Prices below are Q1 2026 list.
| Model | Input ($/M tokens) | Output ($/M tokens) | Best for |
|---|---|---|---|
| Gemini 1.5 Pro | $1.88 | $7.50 | Complex reasoning, document processing, code generation |
| Gemini 1.5 Flash | $0.075 | $0.30 | High-volume, low-latency tasks — service automation, recommendations |
Token metering is where budgets break. A 10,000-token request to Gemini 1.5 Pro costs about $0.02 in input alone; at production volume that compounds fast. Vertex’s permissive default quotas allow burst usage, so configure client-side rate limiting, CUDs and monthly budget alerts before go-live — not after the first bill.
04 Cost at scale
The effective per-seat rate on the Workspace add-on falls with committed volume — but only if you negotiate for it. Google does not publish these bands. Effective monthly rate per user:
The volume-discount conversation gets a senior Google deal team at roughly 1,000 named users — about $360,000 in annual add-on spend. Below that, expect list price unless the add-on rides a Cloud commitment.
05 The Workspace bundling lever
If you take one tactic from this note, take this one. Google discounts the Workspace add-on far more readily when it is attached to Cloud spend than when it stands alone — and the field team’s discretion is widest inside a single, multi-year commercial conversation.
Bundle Cloud commitment + Workspace add-on. Google’s field teams have discretion to cut the $30 add-on to $24–$27 when it rides a three-year Cloud commitment of $150,000+. Framed as one “we’ll sign today” package — with Microsoft Copilot as a live alternative — it lands roughly 65% of the time. Insisting on a single joint Workspace-plus-Cloud negotiation, rather than the two separate ones Google’s reps prefer, is the highest-value move for Cloud-active buyers.
Where an organisation already licenses Vertex AI, push harder still: make Google justify a per-seat add-on that duplicates API access you already pay for. That challenge frequently yields a waived or heavily discounted add-on tier.
06 Buying framework
Four considerations should shape any Gemini commitment. Weight them to your footprint before signing an order.
Base-tier eligibility
Audit that intended users are Business Standard or higher. Business Basic ($6) is not Gemini-eligible, and forced upgrades can quietly add ~50% to the budgeted add-on spend.
Workspace–Cloud overlap
Pull your Vertex AI spend report and compare token consumption against per-seat demand. Existing API usage often makes the per-seat add-on redundant for parts of the estate.
Commitment leverage
A three-year Cloud commitment ($150,000+) is what moves the add-on to $24–$27. Running a Microsoft Copilot alternative in parallel sharpens Google’s pencil further.
Consumption governance
Cap Vertex spend with CUDs and budget alerts, and negotiate 30–90 days’ notice on material model-version changes for any workload you run in production.
07 Our recommendation
Tie the add-on to a three-year Cloud commitment and negotiate them as one deal. Target $24–$27 per seat and a 15–25% blended discount across Workspace, Cloud and security.
Where Vertex AI already covers the workload, skip the per-seat add-on rather than double-pay. Govern consumption with capped expenditure, CUDs and budget alerts.
Don’t buy fleet-wide before a tier audit. Stage licences, lock a 10-business-day activation SLA, and insert model-stability notice language into the MSA.
08 Negotiation sequencing
The process choice matters as much as the terms. How you sequence the two planes decides whether the bundling lever is available at all:
Concurrent Recommended
Negotiate the Workspace add-on and the Cloud commitment as one conversation, with Microsoft Copilot as a live alternative. This is what unlocks $24–$27 seats and the 15–25% blended discount.
Sequential Weaker
Buying the add-on first, then the Cloud commitment, forfeits the bundling lever entirely. Standalone add-on pricing caps you at a 5–10% reduction off list.
Negotiate Gemini before you sign
Our AI procurement practice audits your Workspace and Cloud spend, flags overlap, and benchmarks the add-on against Copilot and market rates.
The Licensing Edge
Weekly AI and licensing intelligence for enterprise IT leaders. 3,000+ subscribers.