A blended rate for human and AI work is a single billing rate that combines the costs of both human staff and AI agents into one figure charged to clients. Firms calculate it as a weighted average of human cost per hour and AI agent cost per billable hour equivalent, then apply a target margin.
When a project is delivered by a team of one human and several AI agents, charging separate rates for each component creates invoice complexity, invites negotiation, and requires more disclosure than most firms are ready to provide. A blended rate solves this. It is a single, defensible number that covers the full delivery team — and it is increasingly how professional services firms are pricing AI-augmented work.
What Is a Blended Rate and Why Does It Matter for AI Work?
The traditional blended rate concept is familiar in professional services: a weighted average of different staff seniority levels into one billing rate. It simplifies invoicing when a project involves partners, managers, and analysts in varying proportions.
The AI-era blended rate extends this concept to include AI agent costs. Instead of weighting only across staff seniority, firms weight across human staff and AI agents by their relative cost contribution to the delivered work.
The key advantage is clarity. Clients want to pay for outcomes and effort. They do not want to audit your AI infrastructure. A blended rate presents the total delivery team as one line on an invoice — the same way a traditional blended rate presents the human team.
There is also a competitive positioning argument. Firms that have demonstrably improved delivery speed and consistency through AI agents should command a blended rate that reflects the value delivered — not a discounted rate that treats AI as a cost reduction to be passed through.
Understanding Your Current Cost Mix
Before setting a blended rate, you need to know what proportion of your delivery costs are human versus AI agent.
Calculate human cost contribution per project: sum the tracked hours per team member multiplied by their fully loaded hourly cost rate. This is already available to firms that track time.
Calculate AI agent cost contribution per project: sum all AI costs attributable to the project — token spend, compute, API fees, oversight time, and a proportional share of fixed AI infrastructure costs. This requires per-project AI cost tracking.
The cost mix ratio is AI agent cost as a percentage of total delivery cost (human + AI). For a typical professional services firm in the early stages of AI adoption, this might be 5–15%. For firms with mature AI pipelines handling research, document processing, and code generation, it may be 25–40% on certain project types.
The mix shifts as AI adoption increases. Setting a blended rate without knowing the current mix — or updating it as the mix changes — produces a rate that is either too low (subsidising clients with your AI investment) or too high (overcharging relative to actual cost).
Track the cost mix per project type quarterly. Different service lines will have very different mixes. A research-heavy consulting engagement might be 35% AI. A client workshop facilitation project might be 3% AI. The blended rate should be set per service line, not globally.
For per-project cost tracking methodology, see AI Agent Cost Tracking for Professional Services and AI Agent Cost Per Billable Hour.
The Blended Rate Formula
Setting the blended rate is a four-step calculation:
Step 1 — Human cost per billable hour. Take the fully loaded annual cost per relevant staff tier, divide by annual productive billable hours. For a mid-level professional at £70,000 fully loaded cost and 1,264 billable hours, this is £55/hr.
Step 2 — AI agent cost per billable hour equivalent. Calculate as described in the AI agent cost per billable hour framework: total monthly AI costs divided by monthly billable hours equivalent produced. For a well-run research agent, this is typically £3–£10/hr equivalent.
Step 3 — Weighted blended cost. Apply the cost mix ratio.
Blended cost per hour = (Human cost/hr × Human weight) + (AI cost/hr × AI weight)Worked example:
- Human cost: £55/hr, weight: 75% (human cost is 75% of total delivery cost)
- AI agent cost: £6/hr equivalent, weight: 25%
- Blended cost: (£55 × 0.75) + (£6 × 0.25) = £41.25 + £1.50 = £42.75/hr
Step 4 — Apply target margin.
Blended billing rate = Blended cost ÷ (1 − target margin)At a 35% target margin: £42.75 ÷ 0.65 = £65.77/hr
This is the rate to charge clients for AI-augmented work delivered by this team composition. Compare it to the human-only rate (£55 ÷ 0.65 = £84.62/hr) — the blended rate is lower because AI has reduced the cost base, but the margin is maintained.
How Do You Communicate a Blended Rate to Clients?
Most clients will not ask how the blended rate is constructed. They want to know what they will pay and what they will receive. For those clients, present the blended rate as a single fee and focus the conversation on outcomes.
When clients ask about AI: prepare a brief, confident explanation that positions AI as a capability, not a cost reduction. A suggested framing: “Our delivery team combines specialist human expertise with AI tools. Our rate reflects the full capability of that team, including improved delivery speed and consistency.”
Contract language should acknowledge AI tool use without becoming a liability disclosure. A clause stating that work may be delivered using AI-assisted tools — and that client data will not be used to train external models — is typically sufficient. This is increasingly standard in professional services contracts.
Avoid itemising AI tool costs in client invoices. Once AI is a named line item, clients will negotiate it as a software cost rather than a delivery capability. The blended rate prevents this framing.
Where transparency is explicitly required — some regulated clients and procurement processes request it — prepare a brief explainer document, not an API invoice. Show methodology, not spend.
How Do You Adjust Blended Rates Over Time?
The blended rate is not a set-and-forget figure. Three forces change it over time.
AI cost deflation. Token prices have fallen consistently. As AI infrastructure costs decrease, the AI component of the blended cost falls. Firms can choose to pass some savings to clients (reducing the blended rate), retain savings as improved margin, or both. Update the formula annually at minimum.
Changing cost mix. As AI adoption increases within the firm, the AI cost weight in the formula grows. A 25% AI cost weight today may be 40% in 18 months. The blended rate must be recalculated when the mix shifts materially (more than 5 percentage points).
Client repositioning risk. If clients begin to perceive AI as making your work cheaper, they may push for rate reductions. Pre-empt this by positioning AI-augmented delivery on value — faster turnaround, greater consistency, expanded capacity — rather than cost. The blended rate conversation is a positioning conversation as much as a pricing calculation.
Review blended rates quarterly. Recalculate the formula when AI costs shift, when the cost mix changes, or when a new service line is introduced. For ongoing ROI tracking, see AI Agent ROI for Professional Services.
Key Takeaway
Blended billing rate = (human cost/hr × human weight + AI cost/hr × AI weight) ÷ (1 − target margin). Recalculate quarterly as your cost mix and AI infrastructure costs shift.
Ready to Set Your AI Blended Billing Rate?
Keito tracks human and AI agent costs per project, giving you the accurate cost mix data you need to set and adjust blended rates with confidence.
Frequently Asked Questions
What is a blended rate for human and AI work?
A blended rate for human and AI work is a single billing rate that combines the fully loaded cost of human staff and the cost-per-billable-hour of AI agents, weighted by their relative contribution to delivery costs, then marked up to the firm’s target margin. It simplifies invoicing, avoids line-item AI cost negotiation, and positions AI-augmented delivery as a unified capability.
How do you calculate a blended billing rate for AI-assisted work?
Four steps: (1) Calculate human cost per billable hour (fully loaded annual cost ÷ annual productive hours). (2) Calculate AI agent cost per billable hour equivalent (total monthly AI costs ÷ monthly billable hours equivalent produced). (3) Apply a weighted average: (human cost/hr × human cost weight) + (AI cost/hr × AI cost weight). (4) Divide by (1 minus your target margin percentage) to get the billing rate.
Should you charge clients separately for AI and human work?
In most cases, no. A blended rate is simpler to manage, avoids AI cost negotiation, and positions AI as a delivery capability rather than a software overhead. Separate billing only makes sense in contexts where clients specifically require itemised AI cost disclosure — typically certain regulated industries or public sector procurement. Even then, use a blended rate as the primary charge with a methodology note rather than itemised API invoices.
How do you communicate a blended rate to clients?
Present the blended rate as a single fee and focus the conversation on what the client receives: faster delivery, greater consistency, and the full capability of your team. If asked about AI, say that your team uses AI tools as part of delivery, and your rate reflects the full delivery capability. Include contract language acknowledging AI tool use without itemising costs.
How often should you adjust your blended rate as AI adoption increases?
Review quarterly. Recalculate the formula when: AI infrastructure costs shift materially (token price changes, new tooling costs), the AI cost mix changes by more than 5 percentage points, or a new service line is introduced. AI token prices are falling consistently — annual recalculation at minimum, quarterly for firms with rapidly evolving AI deployment.