Forge Execution Economics

Overview

Forge Pool uses an execution-native economic model designed for deterministic, replayable, and auditable computation at planetary scale.

Forge pricing is not based on renting idle infrastructure.

Forge pricing derives from:

• verified execution
• deterministic workload attribution
• replay-aware accounting
• finalized execution settlement
• workload-native execution semantics

This allows Forge Pool to provide unusually efficient execution economics across probabilistic simulation, graph propagation, search/discovery, ensemble aggregation, and replay-compatible distributed computation.

---

Pay for Verified Execution — Not Idle Infrastructure

Traditional infrastructure platforms charge for:

• reserved machines
• idle capacity
• opaque infrastructure allocation
• region-specific topology
• nominal hardware ownership

Forge Pool instead charges for:

• verified computation
• finalized execution
• replay-compatible workload attribution
• deterministic execution contribution

This keeps pricing aligned with useful execution rather than infrastructure complexity.

---

The Forge Execution Lifecycle

Every Forge workload passes through a canonical deterministic lifecycle:

request
→ planning
→ execution
→ verification
→ aggregation
→ settlement
→ replayable ledger record

Only finalized execution contributes to:

• billing
• provider payouts
• replay-linked accounting
• execution settlement

This creates a replay-compatible execution economy where accounting emerges directly from computation itself.

---

Why Forge Achieves Efficient Execution Economics

Forge Pool is designed as a distributed planetary execution runtime.

The system achieves efficient economics through:

• heterogeneous distributed execution
• workload-aware orchestration
• deterministic shard execution
• replay-aware settlement
• community-backed provider participation
• infrastructure-neutral execution accounting
• primitive-native workload pricing

Forge scales through contributed execution capacity rather than centralized hyperscale ownership.

---

Primitive-Native Pricing

Forge pricing is workload-native.

Different workload classes have different execution characteristics:

• probabilistic iteration density
• aggregation complexity
• graph propagation cost
• replay requirements
• artifact generation weight
• verification overhead
• memory and execution semantics

Pricing therefore derives from execution semantics rather than generic infrastructure rental.

---

Monte Carlo Execution (mc@1)

Monte Carlo primitives power probabilistic simulation across:

• finance
• insurance
• climate
• weather
• catastrophe modeling
• uncertainty exploration

Indicative Economics

Workload Class	Example Profiles	Indicative Pricing
Financial Simulation	finance.portfolio.v1	from €0.02 / 1M iterations
Insurance Risk	insurance.loss.v1	from €0.03 / 1M iterations
Climate Simulation	climate.ensemble.v1	from €0.03 / 1M iterations
Catastrophe Modeling	insurance.catastrophe.loss_surface.v1	from €0.045 / 1M iterations
Reinsurance Programs	reinsurance.program_loss.v1	from €0.05 / 1M iterations
Weather Stress Surfaces	weather.wind_gust.stress_surface.v1	from €0.04 / 1M iterations

Typical Runtime Characteristics

Illustrative workloads may execute:

• millions of iterations
• across distributed agents
• in tens to hundreds of milliseconds
• with deterministic replay support
• with replay-linked accounting records

---

Graph Propagation Execution (graph@1)

Graph primitives support:

• contagion analysis
• dependency topology
• infrastructure propagation
• cascade modeling
• resilience analysis

Indicative Economics

Workload Class	Example Profiles	Indicative Pricing
Financial Contagion	financial.contagion.v1	from €0.0025 / 1k propagations
Dependency Topology	finance.network.dependence_topology.v1	from €0.002 / 1k graph executions
Infrastructure Propagation	infrastructure.failure.propagation.v1	usage-scaled
Supply Chain Cascade	supply.chain.cascade.v1	usage-scaled
Graph Fragility Analysis	graph.topology.cluster_fragility.v1	usage-scaled

---

Search & Discovery Execution (search@1)

Search primitives support:

• edge-case discovery
• stress mutation
• adversarial scenario exploration
• anomaly ranking
• scenario-space discovery

Indicative Economics

Workload Class	Example Profiles	Indicative Pricing
Stress Discovery	system.stress.v1	from €0.002 / 1k mutations
Tail Scenario Search	insurance.cat.tail.search.v1	usage-scaled
Climate Event Discovery	climate.tail.event.discovery.v1	usage-scaled
Infrastructure Edge Cases	infrastructure.edgecase.discovery.v1	usage-scaled
Financial Regime Search	finance.tail.regime.discovery.v1	usage-scaled

---

Ensemble Aggregation (ensemble@1)

Ensemble primitives support:

• consensus analytics
• risk aggregation
• evidence fusion
• confidence estimation
• multi-model execution surfaces

Indicative Economics

Workload Class	Example Profiles	Indicative Pricing
Risk Aggregation	risk.aggregate.v1	from €0.0015 / 1k aggregations
Climate Consensus	climate.consensus.v1	from €0.0018 / 1k aggregations
Evidence Fusion	evidence.fusion.v1	from €0.0018 / 1k aggregations
Regime Consensus	finance.regime.consensus.v1	usage-scaled
Multi-Model Market Surfaces	finance.multi_model.market_surface.v1	usage-scaled

---

Example Execution Economics

Portfolio Simulation

Example Workload	Runtime	Indicative Cost
1,000,000 portfolio simulations	tens to hundreds of milliseconds	~€0.02
10,000,000 portfolio simulations	distributed execution	~€0.20
100,000,000 portfolio simulations	replayable workload	~€2.00

---

Insurance Scenario Modeling

Example Workload	Runtime	Indicative Cost
1,000,000 insurance simulations	replay-aware execution	~€0.03
10,000,000 catastrophe simulations	distributed execution mesh	~€0.45
Reinsurance aggregation workflow	replay-linked settlement	usage-scaled

---

Graph Contagion Analysis

Example Workload	Runtime	Indicative Cost
10,000 propagation paths	milliseconds	~€0.025
Dependency fragility analysis	replayable graph execution	usage-scaled
Infrastructure cascade exploration	distributed graph execution	usage-scaled

---

Credits and Billing

Forge Pool uses Credits as the normalized unit of verified execution.

Credits are:

• infrastructure-neutral
• replay-compatible
• workload-aware
• settlement-compatible
• deterministic

Credits allow organizations to reason about execution usage independently from:

• hardware ownership
• provider topology
• machine class
• infrastructure region
• execution substrate composition

See also:

• Credits
• Billing

---

Replay-Aware Accounting

Replayability is part of the Forge economic trust model.

Replay support enables organizations to:

• validate historical execution
• reproduce settlement conditions
• audit workload attribution
• verify execution correctness
• investigate anomalies

Replay-linked accounting allows Forge Pool to couple:

• execution
• accounting
• attribution
• settlement
• auditability

into one deterministic runtime model.

See also:

• Execution Accounting
• Economic Integrity

---

Provider Participation

Forge Pool supports a distributed provider economy.

Providers contribute verified execution capacity to the planetary runtime.

Providers may include:

• independent operators
• enterprise fleets
• research institutions
• universities
• infrastructure partners
• hybrid enterprise operators

Provider payouts derive from:

• verified execution contribution
• replay-compatible settlement
• deterministic attribution
• finalized accounting records

Illustrative provider split models may allocate:

• up to 70% to providers
• remaining settlement share to platform orchestration and runtime coordination

Forge rewards useful execution contribution rather than passive infrastructure ownership.

See also:

• Node Payouts
• Provider Policy

---

Enterprise Controls

Forge Pool supports enterprise-grade operational and financial governance controls.

Illustrative controls may include:

• usage caps
• organizational billing scopes
• replay-linked accounting exports
• execution visibility
• project-level attribution
• workload tagging
• provider attribution summaries
• settlement exports

This enables:

• budgeting
• forecasting
• reconciliation
• procurement review
• operational auditability

---

Sandbox and Evaluation Usage

Forge Pool supports sandbox and evaluation execution environments.

Sandbox execution may provide:

• test execution
• limited-scale workloads
• replay experimentation
• Studio workflow testing
• MCP experimentation
• adapter validation

Sandbox usage remains logically separated from production settlement.

---

Marketplace and Ecosystem Participation

Forge Pool supports an execution-native marketplace model.

The ecosystem may include:

• customers
• providers
• adapter publishers
• enterprise operators
• ecosystem partners

Execution remains canonical even as the ecosystem expands.

Marketplace participation remains coupled to:

• replay semantics
• deterministic accounting
• provider attribution
• execution settlement
• governance systems

See also:

• Marketplace
• Adapter Governance

---

What Makes Forge Different

Capability	Traditional Cloud Platforms	Forge Pool
Infrastructure rental	Native	Secondary
Verified execution billing	Rare	Native
Replay-aware accounting	Rare	Native
Deterministic attribution	Limited	Built-in
Distribution-first workloads	Rare	Native
Provider execution economy	Limited	Native
Primitive-native pricing	Rare	Native
Replay-linked settlement	Rare	Native

---

Final Statement

Forge Pool pricing derives from verified execution — not rented infrastructure.

The result is a replay-compatible execution economy designed for:

• probabilistic computation
• distributed execution
• deterministic replay
• provider participation
• enterprise governance
• planetary-scale workload orchestration

Forge Pool does not bill for idle infrastructure complexity.

It bills for useful, attributable, and replay-compatible computation across the planetary runtime.

Execution economics apply consistently across APIs, Studio workflows, adapters, and MCP-compatible agent execution.