Pricing Variables — One Agentic

Section A

LLM API Pricing

Raw model cost per 1M tokens — the inputs everything else derives from.

Parameter	Type	Value	Unit	Notes
Model name	given			Displayed in pricing table notes (e.g. Claude Sonnet, GPT-4o)
Input token price	given		$ / 1M tokens	— — standard prompt & context
Cached input token price	given		$ / 1M tokens	Prompt caching — reused system prompts & templates
Output token price	given		$ / 1M tokens	— — generated diligence response
Revenue markup multiplier	assumption		×	Applied to raw API cost to set customer price
Implied gross margin		—	%	Formula: 1 − (1 ÷ markup)

Section A2

Third-Party API Pricing

External API costs used in session workflows, tracked separately from LLM token costs.

API	Type	Price	Unit	Notes
Tavily	given		$ / credit	Web search API — per-query credit model

Section B

Session Types and Costs

Token counts for both session types. The max uplift % applies to the blended cost and drives the optimistic TAM.

Max Session Uplift

Parameter	Type	Value
Max session uplift	assumption	%

Session Costs & Blended Rate

Session / Token Type	Type	Tokens	Cost
Standard Session
Input	assumption		—
Cached input	assumption		—
Output	assumption		—
Total — Standard		—	— (at markup: —)
Deep Research Session
Input	assumption		—
Cached input	assumption		—
Output	assumption		—
Total — Deep Research		—	— (at markup: —)
Blended Cost
Standard mix	assumption	%	—
Deep research mix (—%)		—	—
Blended cost — base			—
Deep max (—)			—
Blended cost — max (—)			—

Credit Model

Credit Unit & Standard Session

Credit unit is the billing reference — 1 credit = creditUnit tokens of compute. creditRaw is the cost floor per credit; set so the average Neo Interactions session consumes ~3.9 credits at current LLM prices.

Credit Unit

Parameter	Type	Value	Notes
Credit unit size	assumption	tokens	Reference token count per credit
Credit raw cost	assumption		$ per credit — API cost floor
Credit at markup		—	creditRaw × markup

Standard Session — Neo Interactions (avg observed)

Token Type	Type	Cost
Fresh input	observed	—
Cached input	observed	—
Output	observed	—
Avg session cost		—
Credits per avg session		—

Section D

Usage & Spend Estimates

Workflows per user type per month. Fund spend is derived from team composition × per-user rates.

Per User — Monthly Sessions

User Type	Type	Std	Deep	Mo. spend	Yr. spend
Heavy analyst Daily screener, 40+ deals/mo	assumption	—	—	—	—
Moderate analyst Active but not full-time screening	assumption	—	—	—	—
Principal Senior deal lead, strategic evaluation	assumption	—	—	—	—
Partner Spot checks, gut-verify signals	assumption	—	—	—	—

Fund Team Composition

Fund Profile	Heavy Analysts	Moderate Analysts	Principals	Partners	Total users
Small					—
Average ★					—
Active					—
High-volume					—

Per Fund — Monthly & Yearly

Fund Profile	Sessions/mo	Monthly spend (base)	Yearly spend (base)

★ Average fund profile used as basis for competitive comparison and TAM estimates. Workflow counts are derived from team composition × per-user rates.

Section F

TAM — Early-Stage VC Segment

Base uses Average★ fund spend at the base blended session cost. Optimistic applies the max session uplift % to the higher fund count.

Scenario	Type	Active funds	Avg yearly spend	ARR
Base ★	assumption		—	—
Optimistic (—)	assumption		—	—

★ Base case. Base ARPU = Average★ fund spend at base session cost. Optimistic ARPU = Average★ fund spend at max session cost (base × (1 + uplift%)).

Section G

TAM — Angel Investor Segment

Solo angel investors as a separately addressable segment. ARPU is computed from workflow count × blended cost — not a fixed assumption.

Angel Investor Profile

Parameter	Type	Value	Unit	Notes
Monthly workflows per angel Own screening & research, no analyst team	assumption		workflows/mo	Adjust to hit desired ARPU target
Angel ARPU — monthly		—	$/mo	= workflows × blended cost per workflow
Angel ARPU — yearly		—	$/yr	= monthly × 12

Angel Investor TAM (US)

Scenario	Type	ARR
Conservative	assumption	—
Base ★	assumption	—
Optimistic	assumption	—

Angel count basis: Conservative ~30K realistic early adopters; Base 66K = ACA active angel estimate; Optimistic 150K = broader active angel universe. All use the same computed ARPU above.

Combined TAM — VC + Angel

Scenario	VC ARR	Angel ARR	Combined ARR
Base ★	—	—	—
Optimistic	—	—	—

Multi-stage funds, PE, family offices, and corporate VC not included — see Section H below for excluded segment estimates.

Section H

TAM — Excluded Segments (Ballpark)

Multi-stage VC, PE, family offices, and corporate VC. Each segment ARPU = Average VC Fund Annual Spend × Usage Multiplier. All inputs are editable; ARPU and TAM cells are fully derived.

Firm Counts & Multipliers

Segment	Type	Notes
Multi-stage & larger VC	assumption	More deals, larger teams; near-term expansion
Private equity (venture/growth arms)	assumption	Intensive diligence; longer sales cycles
Family offices (direct deals)	assumption	Parity with early-stage VC (conservative)
Corporate VC units	assumption	Moderate deal flow; slow procurement

Derived ARPU & Segment TAM

Segment	Type	Implied ARPU/yr	Firm Count	Segment TAM
Multi-stage & larger VC	derived	—	—	—
Private equity	derived	—	—	—
Family offices	derived	—	—	—
Corporate VC	derived	—	—	—
Total — Excluded Segments				—

Grand Total — All Segments (Base Case)

Component	TAM	Notes
Early-stage VC (base)	—	Primary model — Average★ fund profile
Angel investors (base)	—	Primary model — 66k angels
Excluded segments	—	Ballpark — see Section H above
Grand Total (Base Case)	—	Theoretical ceiling, not near-term forecast

Grand Total — All Segments (Optimistic Case)

Component	TAM	Notes
Early-stage VC (optimistic)	—	Higher fund count × max session cost (+—%)
Angel investors (optimistic)	—	Broader angel universe × max session cost
Excluded segments	—	Same ballpark as base — no separate optimistic variant
Grand Total (Optimistic Case)	—	Theoretical ceiling, not near-term forecast

ARPU multipliers are the key levers here. Adjust them as real-world usage data from analogous segments becomes available. See Excluded Segments Appendix for full detail.