Pricing Strategy: Legionis

The AI-Powered Operating System for Product Organizations

Version: 2.0 Date: 2026-02-18 Owner: BizOps Status: Accepted (Token Bank model approved) Product: Legionis Related: DR-2026-001 (Pricing Model Pivot), DR-2026-004 (Token Bank Pricing), pricing-usage-based-model.md, financial-plan.md, token-economics-byok-vs-managed.md

1. Executive Summary

Pricing Philosophy

Legionis pricing must achieve three objectives simultaneously:

Strategic Intent: We are pricing a Product Operating System, not a chatbot or AI feature. Our pricing should reflect the value category of professional PM tools ($50-150/mo), not consumer AI tools ($10-20/mo).

Key Decisions

Pricing Evolution Context

This strategy reconciles two existing pricing frameworks in the Legionis documentation:

• Financial Plan (v1.0): Feature-gated model at $79/$129 with skill caps (30/500/unlimited)
• Usage-Based Pricing Model (v1.0): Operation-based model at $39/$119 with operation caps

Managed Token Billing: Token Bank Model (v2.0)

As of DR-2026-004 (2026-02-18), managed token billing uses Prepaid Token Banks at 15% markup, replacing the previously proposed 30% metered model. BYOT (Bring Your Own Token/Keys) remains the hero path.

Token Bank Structure:

Key mechanics:

• Prepaid, not metered: Users buy credit packs upfront. No surprise bills, no daily charges.
• Dollar-denominated: Credits are real dollars, never fictional "tokens" or "coins." $1 credit = $1 of Anthropic API cost.
• No expiration at launch: Purchased credits never expire. Revisit after 12 months of usage data.
• Auto-top-up (optional): Users can set a threshold (e.g., "top up $25 when balance drops below $5"). User-configured, never forced.
• Per-task cost preview: Before executing any operation, show estimated cost (e.g., "This PRD will cost ~$0.12"). Users approve before spending.
• Tiered bonuses: $5 bonus at $50 pack, $15 bonus at $100 pack. Encourages larger purchases, improves unit economics.

BYOT remains the hero:

• Messaging: "Bring your own keys for zero markup, or use our credits for instant setup"
• BYOT users get identical platform features; the only difference is who pays for API calls
• Token Banks are positioned as the convenience on-ramp, not the default path

2. Market Context

Pricing Landscape Analysis

The AI/PM tool market creates a pricing paradox: consumer AI tools have established a $10-20/mo anchor, while professional PM tools command $50-150+/mo. Legionis must navigate between these anchors.

Consumer AI Tools (Low Anchor)

Buyer psychology: "AI should cost ~$20/mo" -- this is the anchor we must overcome.

Professional PM Tools (High Anchor)

Buyer psychology: "Professional PM tools cost $50-100+/mo" -- this is the anchor that justifies our pricing.

Our Position

Legionis sits in the gap between these categories. We are not a chatbot with memory (ChatGPT). We are not a legacy roadmapping tool (Productboard). We are a Product Operating System -- a new category.

Price positioning: Above consumer AI ($20), below enterprise PM platforms ($60+), at the professional individual tool price point ($39-119).

Willingness-to-Pay Analysis Framework

Since we have no WTP data yet, here is the framework for validation:

Van Westendorp Price Sensitivity Meter (execute during POC):

Gabor-Granger Direct Questioning (execute during POC):

• "Would you pay $79/mo?" "Would you pay $49/mo?" "Would you pay $39/mo?" "Would you pay $29/mo?"
• Track acceptance rates at each price point

• Test feature bundles vs price sensitivity
• Key dimensions: operations cap, model access, context retention, support level

• Ask 5 beta users all three methods
• Correlate WTP with usage intensity
• Decision: If WTP consistently >$49, hold at $39 (leaves room for increase). If WTP clusters at $29-39, validate at $39 with discount offers.

Category Positioning Decision

Decision: Position as "Product Operating System" -- this justifies pricing above consumer AI tools without requiring enterprise sales cycles. The category creation effort is the marketing cost; the pricing benefit is worth it.

3. Value Metric Analysis

Why "Operations" as the Value Metric

An "operation" is defined as: a single skill invocation, agent spawn, or gateway session that consumes Claude API tokens.

Value metric evaluation criteria:

Score: Operations win on 4 of 6 criteria, particularly on cost alignment and scaling with value.

Alternative Value Metrics Considered

Per-Seat Pricing

• Pros: Simple, predictable, standard SaaS model
• Cons: Does not align with COGS (a user doing 5 ops/day costs us very differently from one doing 200). No natural upgrade trigger. Penalizes team expansion.
• Verdict: Rejected for individual tiers. May be appropriate for Team/Enterprise tiers combined with operation pooling.

Per-Project Pricing

• Pros: Intuitive for product managers ("I manage 3 products")
• Cons: "Project" is ambiguous. Doesn't align with API costs. Hard to enforce.
• Verdict: Rejected. Too vague to build a pricing model around.

Context Storage

• Pros: Measurable, creates lock-in
• Cons: Storage is cheap (<$0.05/GB/mo). Value isn't in storing context -- it's in using it. Would need very low storage limits to create meaningful tiers.
• Verdict: Used as secondary tier differentiator (50MB/500MB/5GB), not primary value metric.

Feature Gating

• Pros: Clear differentiation ("Pro gets X, Free doesn't")
• Cons: Arbitrary -- which skills are "Pro"? All 61 skills deliver value. Gating skills reduces Free tier utility and conversion triggers.
• Verdict: Rejected as primary metric. Used for model access gating (Haiku/Sonnet/Opus) which naturally aligns with quality/cost.

How the Value Metric Aligns with Customer Value Perception

The core insight: an operation is a unit of product work done.

• /prd invocation = a PRD drafted (value: hours saved)
• @pm agent spawn = a PM colleague consulted (value: expertise accessed)
• @plt gateway = a leadership team meeting run (value: cross-functional alignment)

• Users can intuitively see that more operations = more value received
• The ROI display creates a natural "this paid for itself" moment
• Hitting the cap is a positive signal ("I'm getting value") not a negative one

4. Tier Architecture

Tier Overview

*Fair use policy: 50K ops/month soft cap for Pro Plus

Free Tier: The Evaluation Engine

Target Persona: Individual PMs evaluating the product, hobbyists, open-source plugin users curious about the SaaS experience.

What They Get:

• 500 operations/month (~17 ops/day)
• Access to all 61 skills (no skill gating)
• Access to all 13 agent personas (inline mode only)
• Haiku model only (cost containment)
• 7-day context retention
• 50MB storage
• Community support
• ROI tracking (the conversion engine)

• Sonnet and Opus models (quality upgrade)
• Extended context retention (meaningful work requires >7 days)
• Storage beyond 50MB
• Priority queuing
• Direct support

Pricing Rationale: Free tier is a pure acquisition cost. Philosophy: enable meaningful evaluation without creating COGS liability. The 500-op cap allows ~2-3 weeks of moderate daily use before hitting limits. Haiku-only gating is the primary COGS control.

COGS Estimate:

• Avg usage: [TBD - measure actual, likely 200-400 of 500 cap]
• Model: 100% Haiku
• Estimated COGS: [TBD - calculated from Haiku rate x avg token usage per op]
• Target: <$5/user/month

Pro Tier: The Revenue Engine ($39/mo)

Target Persona: Individual PMs doing regular product work. "Strategic Sarah" -- the senior PM who needs this daily.

What They Get:

• 5,000 operations/month (~167 ops/day)
• Haiku + Sonnet models
• Limited Opus access (3 invocations/day for high-value operations)
• 90-day context retention
• 500MB storage
• Priority queue (faster responses)
• Email support (48hr SLA)
• Overage billing (see Section 4.6)

• Full Opus access (Pro Plus)
• Dedicated capacity (Pro Plus)
• Extended context retention beyond 90 days (Pro Plus)
• Storage beyond 500MB (Pro Plus)

Pricing Rationale: $39 is the compromise position between PMM ($29, accessibility) and BizOps ($49, margin protection):

• Leaves $9 margin cushion if COGS hits $30/user (the critical constraint)
• Competitive with professional tools (not trying to match ChatGPT at $20)
• Psychological: under $40, which is a common SaaS threshold
• Room to increase to $49 if COGS proves manageable, or decrease to $29 if market demands it

• Target model mix: 70% Haiku, 25% Sonnet, 5% Opus
• Avg usage: [TBD - measure actual, likely 2,000-3,500 of 5,000 cap]
• Critical constraint: Pro COGS MUST stay <$30/user/month
• Margin target: >40% ($39 - $30 max COGS = $9 minimum margin = 23% minimum)
• If COGS <$25: comfortable 36%+ margin, consider lowering to $29
• If COGS >$30: must raise to $49 or tighten model routing

Pro Plus Tier: The Power User Tier ($119/mo)

Target Persona: Product leaders managing multiple products, fractional CPOs, consultants serving multiple clients. "Scaling Steve" at an individual level.

What They Get:

• Unlimited operations (fair use: 50K/mo soft cap)
• Full model access: Haiku, Sonnet, and Opus
• Dedicated capacity (guaranteed response times, no queueing)
• 1-year context retention
• 5GB storage
• Priority email + Slack channel support
• Early access to new features

• Team collaboration features (Team tier)
• SSO, admin controls (Enterprise)
• Shared workspaces (Team/Enterprise)

Pricing Rationale: $119 maintains premium positioning:

• 3x Pro pricing (clear value gap -- unlimited ops + full Opus is a massive quality difference)
• "Just under $120" psychological threshold
• High margin even with heavy Opus usage (model mix: 40% Haiku, 40% Sonnet, 20% Opus)
• Pro Plus users self-select for high-value work -- they're doing strategic bets, PLT sessions, portfolio reviews

• Target model mix: 40% Haiku, 40% Sonnet, 20% Opus
• Avg usage: [TBD - measure actual from beta users]
• Target COGS: <$55/user/month (53% margin minimum)
• If heavy Opus users push COGS >$70: review fair use enforcement or adjust Opus routing

• 50K ops/month soft cap
• At 50K: Automated email: "You're a power user! We'd love to hear your use case."
• At 100K: Account review initiated (sales call to understand use case)
• Abuse confirmed (resale, bot usage): Warning, then suspension
• Legitimate power users are never penalized -- enforcement targets abuse only

Team Tier (Future -- Phase 3): Collaborative Product OS ($59/user/mo)

Target Persona: Product teams of 5-15 PMs. "Scaling Steve" buying for his team.

Design Principles:

• Per-seat pricing (standard for team tools)
• Pooled operations (5,000 ops/user/month, pooled across team)
• Shared context registry (decisions, bets, feedback visible to all team members)
• Admin controls (invite/remove users, usage visibility)
• Portfolio views (cross-product context)

• $20/seat discount from Pro ($79 value with collaboration features)
• Minimum 5 seats to filter out "fake teams"
• Competitive with PM tools: cheaper than Productboard ($60-180), more expensive than Notion ($15)
• Expansion revenue: teams start at 5, grow to 10-20 seats

Enterprise Tier (Future -- Phase 5): Custom ($15K+/yr)

Target Persona: Enterprise product organizations with 50+ PMs. "Compliance Carol" -- the CPO.

Design Principles:

• Annual contract value (ACV) based pricing
• Unlimited operations, all models, unlimited storage
• SSO (SAML, Okta, Azure AD)
• Advanced admin controls, audit logs, RBAC
• Dedicated support with SLA (<4hr response)
• Custom deployment options (dedicated instance, data residency)
• Professional services (V2V framework implementation, training)

• CAC for enterprise is $5K-10K (sales cycle, demo, legal, POC)
• Need $150K+ ACV to justify sales investment at scale
• Volume discounts at 50+ seats (30-50% off retail)
• Multi-year discounts (10-20% off for 2-3 year commits)

Overage Billing (Pro Tier Only)

Free tier has a hard cap (no overages). Pro Plus is unlimited. Overages apply only to Pro tier.

Design Intent:

• Tiered pricing creates natural upgrade incentive: at 10K ops, overage = $42, close to Pro Plus's $80 price difference
• Rates are above COGS but below Pro Plus (profitable but not punitive)
• At 80% cap (4,000 ops), show in-app projection: "At this pace, you'll hit 6,200 ops. Estimated overage: $12. Or upgrade to Pro Plus for unlimited."

5. COGS Framework

Cost Structure per Tier

Model Cost Structure (populate with actual Anthropic API rates):

Average Operation Token Usage (Framework)

Measurement Method: Run 100 operations of each type in production environment, log token counts, calculate averages. This MUST be done before launch pricing is finalized.

Expected Model Mix by Tier

Validation Method: Track actual model mix in first 90 days. If Pro tier Opus usage exceeds 5%, tighten daily Opus limit from 3 to 1 or remove entirely.

Blended COGS Calculation Framework

Tier COGS per User = (Avg Operations/User/Month) x (Blended Cost per Operation)
Where:
Blended Cost per Operation =
  (Model Mix % Haiku x Haiku Op Cost) +
  (Model Mix % Sonnet x Sonnet Op Cost) +
  (Model Mix % Opus x Opus Op Cost)

Margin Analysis Framework

SaaS Benchmark: Healthy SaaS companies target 60-80% gross margins. At launch, we accept lower margins (23-36% for Pro) to prioritize accessibility, with a clear path to 60%+ via COGS optimization.

Intelligent Model Routing (COGS Optimization)

Goal: Minimize COGS by routing operations to the least expensive capable model.

Routing Rules (applied automatically):

Complexity Detection Triggers (auto-escalation):

• User prompt >500 words
• Multiple @file context references
• Multi-agent gateway invocation
• Skill requires synthesis of >3 data sources

COGS Optimization Levers (Priority Order)

If actual COGS exceeds targets, apply these levers in order:

Recommendation: Start with levers 1-3 (routing, caching, output limiting) before restricting user access or raising prices. These improve margins without degrading the user experience.

6. Conversion & Retention Strategy

Free --> Pro Conversion

Target: 5-10% conversion rate within 90 days of signup

Primary Conversion Mechanism: ROI Visibility

The built-in ROI tracking is the most powerful conversion tool. Every skill and agent invocation shows time saved:

Completed: PRD for authentication feature
Time saved: ~4 hours (vs. manual PRD writing + stakeholder reviews)
Session total: 12.5 hours saved this month

Conversion prompt at cap or ROI threshold:

"This month you saved 12.5 hours using Product Org OS (500 operations).

Upgrade to Pro for 5,000 ops/month -- that's 100+ hours saved for $39/mo.

That's less than $0.40 per hour saved."

Secondary Conversion Mechanics:

Pro --> Pro Plus Upgrade

Target: 15-20% of Pro users upgrade within 6 months

Primary Upgrade Trigger: Overage Accumulation

When Pro users consistently hit their 5,000 op cap, overages create a natural price comparison:

"Your overages this month: $52 (7,200 operations).

Pro Plus is $119/mo for unlimited operations.

At your usage level, Pro Plus saves you $[overage - $80 price difference]/month."

Secondary Upgrade Triggers:

Retention Through Context Lock-In

The context moat is the primary retention mechanism. As users accumulate decisions, bets, assumptions, learnings, and feedback in their context registry, switching cost increases:

Retention tactics by lifecycle stage:

• Day 1-7: Onboarding -- guide to first decision record, first PRD. Goal: 3+ operations.
• Day 7-30: Habit formation -- weekly "you saved X hours" email. Goal: 50+ operations.
• Day 30-90: Value realization -- show accumulated context value. Goal: 200+ operations, 20+ context items.
• Day 90+: Lock-in -- context becomes indispensable. Goal: user references past decisions regularly.

ROI-Driven Conversion (The Flywheel)

User invokes skill/agent
  --> ROI displayed ("~4 hrs saved")
  --> Session total accumulates
  --> Weekly email: "You saved 15 hours this month"
  --> Monthly: "You've saved 45 hours in 3 months"
  --> At cap: "Upgrade to save even more"
  --> After upgrade: "Since upgrading, you've saved 120 hours ($4,680 value at $39/hr PM rate)"

This creates a self-reinforcing conversion loop: the more the user engages, the more value they see, the stronger the upgrade signal.

7. Pricing Alternatives & Recommendations

Option A: Accessibility Pricing ($29/$99)

Philosophy: Maximize TAM, compete on price, grow fast

Financial Implications Framework:

If 1,000 Pro users, COGS = $28/user:
Revenue = $29,000/mo
COGS = $28,000/mo
Gross Profit = $1,000/mo (3.4% margin)
If COGS climbs to $31/user (10% overage):
Revenue = $29,000/mo
COGS = $31,000/mo
Gross Profit = -$2,000/mo (LOSS)

This option works if:

• COGS stays reliably <$27 (leaves $2 buffer)
• Free-to-Pro conversion rate >15% (volume compensates thin margin)
• Pro Plus adoption >20% of Pro base (high-margin tier subsidizes)
• Prompt caching achieves >60% hit rate

Option B: Margin-First Pricing ($49/$129)

Philosophy: Protect margins, position as premium, grow sustainably

Financial Implications Framework:

If 1,000 Pro users, COGS = $30/user:
Revenue = $49,000/mo
COGS = $30,000/mo
Gross Profit = $19,000/mo (38.8% margin)
If COGS climbs to $35/user (17% overage):
Revenue = $49,000/mo
COGS = $35,000/mo
Gross Profit = $14,000/mo (28.6% margin -- still viable)

This option works if:

• Product delivers clear value vs. consumer AI tools
• Target market is professional/enterprise (not prosumer)
• Premium positioning as "AI product organization" justifies price
• Conversion rate accepts higher price point (likely 3-5% vs 5-10%)

Option C: Hybrid Pricing ($39/$119) -- RECOMMENDED

Philosophy: Balance accessibility with margin protection. Data-driven adjustment post-launch.

Financial Implications Framework:

If 1,000 Pro users, COGS = $25/user (target):
Revenue = $39,000/mo
COGS = $25,000/mo
Gross Profit = $14,000/mo (35.9% margin)
If COGS = $30/user (worst case before increase):
Revenue = $39,000/mo
COGS = $30,000/mo
Gross Profit = $9,000/mo (23.1% margin -- thin but viable)
If COGS = $20/user (post-optimization):
Revenue = $39,000/mo
COGS = $20,000/mo
Gross Profit = $19,000/mo (48.7% margin -- healthy)

Why Hybrid is Recommended:

Comparison Summary

Recommendation: Launch at Option C ($39/$119). Gather 90 days of COGS data and conversion data. Adjust based on evidence, not assumption.

8. Future Pricing Evolution

Team Tier Design (Phase 3 -- Q4 2026)

Pricing Model: Per-seat + shared context pool

Volume Discounts:

• 10+ seats: 10% off ($53/user/mo)
• 25+ seats: 20% off ($47/user/mo) -- qualifies for Enterprise track
• 50+ seats: custom pricing (Enterprise)

• $20 discount vs. Pro ($79 implied value with collaboration features)
• Signal: "Bring your team, save per-seat"
• Competitive: Below Productboard ($60-180), above Notion ($15)
• Minimum 5 seats ($295/mo entry) filters out non-team buyers

Enterprise Tier Design (Phase 5 -- Q2 2027)

Pricing Model: Annual contract value (ACV) based

ACV Frameworks by Company Size:

• 10-25 PMs: $15K-30K/yr
• 25-50 PMs: $30K-75K/yr
• 50-100 PMs: $75K-150K/yr
• 100+ PMs: $150K-500K/yr (includes professional services)

• 50+ seats: 30% off retail
• Multi-year (2yr): 10% off
• Multi-year (3yr): 20% off
• Case study/reference agreement: additional 10-20% off

Price Increase Strategy

Grandfathering Policy: Existing users maintain their price for 12 months after any increase. After 12 months, prices adjust to current rates.

Increase Triggers:

• COGS consistently exceeds 70% of Pro price for 3+ months
• Market WTP data (Van Westendorp/Gabor-Granger) shows headroom
• Significant new capability added (e.g., multi-agent, PLT gateway)
• Competitive landscape shift (competitors raise prices)

International Pricing Considerations

Phase 1 (Launch): Single global price in USD. Simple, no complexity.

Phase 2 (M6-12): Consider purchasing power parity (PPP) for key markets:

• India, Southeast Asia, Eastern Europe: 30-50% discount
• Western Europe, Australia, Japan: Full price (similar purchasing power)
• Implementation: IP-based detection + honor system (company domain verification)

• EUR pricing for EU (GDPR data residency may require EU hosting)
• GBP pricing for UK
• INR pricing for India (PPP adjusted)

9. Implementation Plan

Launch Pricing Decisions Needed (Before Launch)

A/B Testing Plan

Test 1: Price Point (POC Period)

• Segment: 50% of POC users see "$39/mo" messaging, 50% see "$49/mo"
• Metric: "Would you pay this?" response rate
• Duration: 2 weeks
• Decision: If $49 acceptance rate >80% of $39 rate, launch at $49

• Segment: 50% of new Free users get 500 ops cap, 50% get 300 ops
• Metric: Free-to-Pro conversion rate at 30 and 60 days
• Duration: 60 days
• Decision: If 300-cap conversion rate is >2x 500-cap, lower Free cap permanently

• Segment: 3-way test of upgrade CTAs

• Metric: Click-through rate and upgrade conversion rate
• Duration: 30 days per variant
• Decision: Use winning variant as primary upgrade CTA

• Segment: 50% of upgrade page shows "Monthly $39 | Annual $31.20/mo (20% off)", 50% shows "Monthly $39" only
• Metric: Annual adoption rate, net revenue impact
• Duration: 90 days
• Decision: If annual adoption >30%, keep annual option prominently displayed

90-Day Validation Metrics

Track daily during first 90 days:

Adjustment Triggers and Protocol

10. Appendices

A. COGS Calculation Frameworks

Framework 1: Blended COGS per Operation

Input:
Haiku cost per op: $[TBD]
Sonnet cost per op: $[TBD]
Opus cost per op: $[TBD]
Model mix (H/S/O): [70/25/5] for Pro, [40/40/20] for Pro Plus

Output:
Pro Blended Cost = (0.70 x Haiku) + (0.25 x Sonnet) + (0.05 x Opus)
Pro Plus Blended Cost = (0.40 x Haiku) + (0.40 x Sonnet) + (0.20 x Opus)

Framework 2: Monthly COGS per User

Input:
Avg operations/user/month: [TBD]
Blended cost per op: $[calculated above]

Output:
Monthly COGS = Avg Ops x Blended Cost

Framework 3: Margin Sensitivity

At $39 Pro price:
COGS $15/user --> 61.5% margin (excellent)
COGS $20/user --> 48.7% margin (good)
COGS $25/user --> 35.9% margin (acceptable)
COGS $30/user --> 23.1% margin (minimum viable)
COGS $35/user --> 10.3% margin (REQUIRES ACTION)
COGS $40/user --> -2.6% margin (LOSS - raise price or cut costs)

B. Competitive Pricing Benchmarks

Positioning Insight: At $39/mo, we are:

• 2x ChatGPT/Claude Pro (justified by product-specific value + context persistence)
• 4x GitHub Copilot (different category -- product work vs. coding)
• 0.65x Productboard entry (cheaper than PM tools, AI-native)
• 0.66x Aha! entry (cheaper, with AI-native advantages)

C. Pricing Page Wireframe Suggestions

Layout: 3-column comparison (Free | Pro | Pro Plus), with Team and Enterprise as separate section below.

Key Elements per Tier Card:

Psychological Pricing Tactics:

• Show Pro as "Most Popular" (social proof, even if aspirational)
• Show monthly and annual toggle (annual = 20% off)
• Anchor Pro Plus as "Best Value for Power Users"
• Show ROI calculator: "How much will you save?" (input hours/week on product work, output monthly value vs. price)

• Headline: "Start free. Scale as you grow."
• Sub-headline: "Every PM starts with 500 free operations/month. Upgrade when you need more."

D. Revenue Model by Pricing Scenario (Per 10,000 Total Users)

Assumptions: 70% Free, 25% Pro, 5% Pro Plus

Revenue Delta: Option C generates $420K more ARR than Option A, and $360K less than Option B. The $420K additional revenue over Option A is worth the marginal conversion risk.

E. Glossary

Document Control

Version: 2.0 Last Updated: 2026-02-18 Owner: BizOps Next Review: After POC completion (Week 6) -- pricing decisions finalized Distribution: Product Leadership Team, BizOps, PMM

Related Documents:

• pricing-usage-based-model.md -- Operational pricing model (superseded by this strategy)
• financial-plan.md -- Unit economics and financial projections
• strategic-plan.md -- Product strategy and phased roadmap
• gtm-strategy.md -- Go-to-market strategy
• token-economics-byok-vs-managed.md -- BYOK vs managed billing analysis (v2.0)
• decisions/DR-2026-004-token-bank-pricing.md -- Token Bank pricing decision record

• 2026-02-18: v2.0 -- Added Token Bank prepaid model at 15% markup (replaces 30% metered). Added credit pack structure, auto-top-up, no-expiration policy, per-task cost preview. BYOT remains hero path. See DR-2026-004.
• 2026-01-28: v1.0 -- Comprehensive pricing strategy consolidating usage-based model and financial plan insights

End of Document