Cost Projection

@webframp/cost-projectionv2026.09.19.1· 17d agoMODELSREPORTS

01README

GPU inference cost projection across cloud, rental, and capex scenarios.

Three model types store scenario inputs (instance rates, hardware costs, facility expenses) and compute normalized $/GPU-hour projections. A comparison report queries all scenarios and produces a ranked table with crossover analysis.

No live API calls — rates are manually entered from quotes, pricing pages, or account team conversations. All scenarios in a comparison must share a currency.

Model Types

gpu-cloud — Hyperscaler GPU instances (AWS, Azure, GCP) under various capacity models: on-demand, reserved, FTP, Capacity Blocks, etc.
gpu-rental — Third-party providers (CoreWeave, Lambda Labs, RunPod) with simpler per-GPU-hour pricing and optional commitment discounts.
gpu-capex — On-premises/colo hardware. Amortizes a capital purchase into synthetic recurring cost, adds facility/staff/maintenance, and normalizes to $/GPU-hour. Includes sensitivity analysis across utilization and useful-life assumptions.

Usage

swamp extension pull @webframp/cost-projection

# Cloud scenario
swamp model create @webframp/cost-projection/gpu-cloud kimi-k3-hyperpod
swamp model method run kimi-k3-hyperpod record \
  --input name=kimi-k3-hyperpod \
  --input provider=aws \
  --input region=us-east-1 \
  --input instanceType=ml.p6-b300.48xlarge \
  --input gpuCount=8 \
  --input gpuModel="NVIDIA B300" \
  --input capacityModel=flexible-training-plan \
  --input instanceRatePerHour=98.50

# Rental scenario
swamp model create @webframp/cost-projection/gpu-rental kimi-k3-coreweave
swamp model method run kimi-k3-coreweave record \
  --input name=kimi-k3-coreweave \
  --input provider=coreweave \
  --input gpuModel="NVIDIA H100 SXM" \
  --input gpuCount=8 \
  --input ratePerGpuHour=2.49

# Capex scenario
swamp model create @webframp/cost-projection/gpu-capex dc-east-b300
swamp model method run dc-east-b300 record \
  --input name=dc-east-b300 \
  --input gpuModel="NVIDIA B300" \
  --input gpuCount=8 \
  --input gpuCostPerUnit=35000 \
  --input serverCost=45000 \
  --input networkingCost=15000 \
  --input totalHardwareCost=340000 \
  --input usefulLifeMonths=36 \
  --input coloCostPerKwMonth=150 \
  --input powerDrawKw=10.5

# Run sensitivity analysis
swamp model method run dc-east-b300 sensitivity

# The comparison report runs automatically after any method call on any
# gpu-cloud, gpu-rental, or gpu-capex instance. Retrieve its latest output:
swamp report get @webframp/cost-projection-comparison --model dc-east-b300 --markdown

Per-Token Break-Even (Optional)

To compare self-hosted cost against per-token API pricing, provide apiComparisonRatePerMToken ($/million tokens from the API provider) and estimatedTokensPerGpuHour (from vLLM benchmarks or model card throughput figures) when recording a scenario. The projection will compute the monthly token volume at which self-hosting breaks even.

02Release Notes

2026.09.19.1

Changed: Removed the deprecated drivers: manifest key. Swamp core removed extension driver support (swamp-club/swamp#2333, 2026-09-01); the key now emits a deprecation warning and is ignored by the manifest parser. The field was an empty list, so this is a no-op for behavior — the extension ships the same three models and one report as before.

Upgrade note: No migration required. No model schema, method, or resource changed.

03Models3

@webframp/cost-projection/gpu-cloudv2026.09.19.1cost-projection/gpu_cloud.ts

fn record(name: string, region: string, instanceType: string, gpuCount: number, gpuModel: string, commitmentTermMonths?: number, instanceRatePerHour: number, currency: string, hoursPerDay: number, daysPerMonth: number, replicas: number, storageGb: number, storageRatePerGbMonth: number, dataTransferGbMonth: number, dataTransferRatePerGb: number, managementFeePerMonth: number, apiComparisonRatePerMToken?: number, estimatedTokensPerGpuHour?: number, notes?: string, sourceUrl?: string, quotedAt?: string, fetchedAt?: string, durationMs?: number, collectedBy?: string)

Record a cloud GPU inference scenario and compute its projection.

Argument	Type	Required	Description
name	string	yes	Human-readable name for this scenario
region	string	yes	Cloud region for the instance (e.g. us-east-1)
instanceType	string	yes	Cloud provider instance type or SKU
gpuCount	number	yes	Number of GPUs per instance
gpuModel	string	yes	GPU model (e.g. H100, A100)
commitmentTermMonths?	number	no	Length of the commitment term in months, if the capacity model requires one
instanceRatePerHour	number	yes	Quoted hourly rate per instance, in currency units
currency	string	yes	Currency of all rate fields in this scenario
hoursPerDay	number	yes	Expected hours of instance utilization per day
daysPerMonth	number	yes	Expected days of instance utilization per month
replicas	number	yes	Number of identical instances running concurrently
storageGb	number	yes	Attached storage size per instance, in GB
storageRatePerGbMonth	number	yes	Storage cost per GB per month
dataTransferGbMonth	number	yes	Expected monthly data transfer volume, in GB
dataTransferRatePerGb	number	yes	Data transfer cost per GB
managementFeePerMonth	number	yes	Flat monthly management/platform fee, independent of replicas
apiComparisonRatePerMToken?	number	no	Comparable managed-API rate per million tokens, used to compute break-even
estimatedTokensPerGpuHour?	number	no	Estimated throughput in tokens per GPU-hour, from benchmarks or model card figures
notes?	string	no	Free-form notes about this scenario
sourceUrl?	string	no	URL of the quote or pricing page this scenario is based on
quotedAt?	string	no	Date the rate was quoted (YYYY-MM-DD)
fetchedAt?	string	no	ISO 8601 timestamp when data was fetched
durationMs?	number	no	Method execution duration in milliseconds
collectedBy?	string	no	Extension that collected this data

fn project()

Re-compute the projection from the stored scenario. Use after

fn update_rate(instanceRatePerHour: number, quotedAt?: string)

Update the instance hourly rate and quotedAt timestamp without

Argument	Type	Required	Description
instanceRatePerHour	number	yes	New quoted hourly rate per instance, in the scenario's currency
quotedAt?	string	no	Date the new rate was quoted (YYYY-MM-DD); defaults to today

Resources

scenario(infinite)— Cloud GPU inference scenario — instance type, capacity model,

projection(infinite)— Computed cost projection normalized to $/GPU-hour. Derived from

@webframp/cost-projection/gpu-rentalv2026.09.19.1cost-projection/gpu_rental.ts

fn record(name: string, provider: string, region?: string, gpuModel: string, gpuCount: number, gpuMemoryGb?: number, ratePerGpuHour: number, currency: string, commitmentDiscountPct: number, hoursPerDay: number, daysPerMonth: number, storageGb: number, storageRatePerGbMonth: number, networkEgressGbMonth: number, networkEgressRatePerGb: number, apiComparisonRatePerMToken?: number, estimatedTokensPerGpuHour?: number, notes?: string, sourceUrl?: string, quotedAt?: string, fetchedAt?: string, durationMs?: number, collectedBy?: string)

Record a GPU rental scenario and compute its projection.

Argument	Type	Required	Description
name	string	yes	Human-readable name for this scenario
provider	string	yes	GPU rental provider name (e.g. CoreWeave, Lambda Labs)
region?	string	no	Provider region or datacenter location
gpuModel	string	yes	GPU model (e.g. H100, A100)
gpuCount	number	yes	Number of GPUs in this scenario
gpuMemoryGb?	number	no	Memory per GPU, in GB
ratePerGpuHour	number	yes	Quoted list price per GPU per hour
currency	string	yes	Currency of all rate fields in this scenario
commitmentDiscountPct	number	yes	Discount off ratePerGpuHour granted for the commitment term, as a percentage
hoursPerDay	number	yes	Expected hours of GPU utilization per day
daysPerMonth	number	yes	Expected days of GPU utilization per month
storageGb	number	yes	Attached storage size, in GB
storageRatePerGbMonth	number	yes	Storage cost per GB per month
networkEgressGbMonth	number	yes	Expected monthly network egress volume, in GB
networkEgressRatePerGb	number	yes	Network egress cost per GB
apiComparisonRatePerMToken?	number	no	Comparable managed-API rate per million tokens, used to compute break-even
estimatedTokensPerGpuHour?	number	no	Estimated throughput in tokens per GPU-hour, from benchmarks or model card figures
notes?	string	no	Free-form notes about this scenario
sourceUrl?	string	no	URL of the quote or pricing page this scenario is based on
quotedAt?	string	no	Date the rate was quoted (YYYY-MM-DD)
fetchedAt?	string	no	ISO 8601 timestamp when data was fetched
durationMs?	number	no	Method execution duration in milliseconds
collectedBy?	string	no	Extension that collected this data

fn project()

Re-compute the projection from the stored scenario. Use after

fn update_rate(ratePerGpuHour: number, quotedAt?: string)

Update the per-GPU hourly rate and quotedAt without re-entering

Argument	Type	Required	Description
ratePerGpuHour	number	yes	New quoted list price per GPU per hour
quotedAt?	string	no	Date the new rate was quoted (YYYY-MM-DD); defaults to today

Resources

scenario(infinite)— GPU rental inference scenario — provider, GPU type, hourly rate,

projection(infinite)— Computed cost projection normalized to $/GPU-hour. Derived from

@webframp/cost-projection/gpu-capexv2026.09.19.1cost-projection/gpu_capex.ts

fn record(name: string, site?: string, gpuModel: string, gpuCount: number, gpuCostPerUnit: number, serverCost: number, networkingCost: number, totalHardwareCost: number, usefulLifeMonths: number, residualValuePct: number, coloCostPerKwMonth: number, powerDrawKw: number, pue: number, networkBandwidthCostPerMonth: number, staffFteAllocation: number, staffCostPerFteMonth: number, failureRatePctPerYear: number, spareBudgetPerMonth: number, warrantyMonths: number, targetUtilizationPct: number, hoursPerDay: number, daysPerMonth: number, apiComparisonRatePerMToken?: number, estimatedTokensPerGpuHour?: number, currency: string, notes?: string, quotedAt?: string, fetchedAt?: string, durationMs?: number, collectedBy?: string)

Record a capex GPU inference scenario and compute its projection.

Argument	Type	Required	Description
name	string	yes	Human-readable name for this scenario
site?	string	no	Physical site or datacenter location
gpuModel	string	yes	GPU model (e.g. H100, A100)
gpuCount	number	yes	Number of GPUs purchased
gpuCostPerUnit	number	yes	Purchase price per GPU
serverCost	number	yes	Server/chassis hardware cost, excluding GPUs
networkingCost	number	yes	Networking hardware cost (switches, NICs, cabling)
totalHardwareCost	number	yes	Total upfront hardware cost; should equal gpuCostPerUnit * gpuCount + serverCost + networkingCost
usefulLifeMonths	number	yes	Amortization period for the hardware, in months
residualValuePct	number	yes	Expected resale/residual value at end of useful life, as a percentage of totalHardwareCost
coloCostPerKwMonth	number	yes	Colocation/power cost per kW per month
powerDrawKw	number	yes	IT power draw of the hardware, in kW
pue	number	yes	Power usage effectiveness multiplier applied to powerDrawKw to account for cooling/overhead
networkBandwidthCostPerMonth	number	yes	Flat monthly network bandwidth cost
staffFteAllocation	number	yes	Fraction of staff FTE allocated to operating this hardware
staffCostPerFteMonth	number	yes	Fully-loaded staff cost per FTE per month
failureRatePctPerYear	number	yes	Expected annual hardware failure rate, as a percentage of totalHardwareCost, used to estimate replacement spend
spareBudgetPerMonth	number	yes	Flat monthly budget reserved for spare parts
warrantyMonths	number	yes	Manufacturer warranty coverage, in months
targetUtilizationPct	number	yes	Target GPU utilization used to compute the effective cost per GPU-hour
hoursPerDay	number	yes	Expected hours of hardware availability per day
daysPerMonth	number	yes	Expected days of hardware availability per month
apiComparisonRatePerMToken?	number	no	Comparable managed-API rate per million tokens, used to compute break-even
estimatedTokensPerGpuHour?	number	no	Estimated throughput in tokens per GPU-hour, from benchmarks or model card figures
currency	string	yes	Currency of all cost fields in this scenario
notes?	string	no	Free-form notes about this scenario
quotedAt?	string	no	Date the hardware cost was quoted (YYYY-MM-DD)
fetchedAt?	string	no	ISO 8601 timestamp when data was fetched
durationMs?	number	no	Method execution duration in milliseconds
collectedBy?	string	no	Extension that collected this data

fn project()

Re-compute the projection from the stored scenario.

fn update_hardware_cost(totalHardwareCost: number, gpuCostPerUnit?: number, quotedAt?: string)

Revise the total hardware cost (e.g. after a new vendor quote)

Argument	Type	Required	Description
totalHardwareCost	number	yes	New total upfront hardware cost
gpuCostPerUnit?	number	no	New purchase price per GPU; leave unset to keep the stored value
quotedAt?	string	no	Date the new cost was quoted (YYYY-MM-DD); defaults to today

fn sensitivity()

Run the projection across a matrix of utilization and useful-life

Resources

scenario(infinite)— Capex GPU inference scenario — hardware costs, amortization

projection(infinite)— Computed cost projection with amortized hardware normalized to

sensitivity(infinite)— Sensitivity matrix showing $/GPU-hour across multiple utilization

04Reports1

@webframp/cost-projection-comparisonmodel

scenario_comparison.ts

Cross-scenario GPU inference cost comparison, normalized to $/GPU-hour.

gpucostprojectioncomparisonfinops

05Previous Versions12

2026.09.18.1Sep 20, 2026

2026.09.18.1

Upgrade note: Normalized npm:zod dependency version to 4.6.5 across the repo. No behavioral changes in this extension.

2026.09.15.1Sep 15, 2026

2026.09.15.1

Changed: Bump zod 4.4.3 → 4.6.5

2026.08.28.1

Changed: Normalized the extension license to Apache-2.0 and corrected the copyright holder to "Sean Escriva". Extensions that previously shipped an MIT LICENSE.md are now Apache-2.0, consistent with the repository root and every other extension. No code or behavioral changes.

Upgrade note: License text only. No API, schema, or runtime behavior changed.

2026.08.26.3

Fixed: Restored inline npm:zod@4.4.3 import specifiers so the registry quality scorer can resolve dependencies and score the extension. An earlier release used a bare "zod" import-map specifier, which published but scored as unscored.

Changed: Retained explicit compilerOptions.strict in deno.json. No behavioral or schema changes.

2026.08.28.1Aug 28, 2026

2026.08.28.1

Upgrade note: License text only. No API, schema, or runtime behavior changed.

2026.08.26.3

Changed: Retained explicit compilerOptions.strict in deno.json. No behavioral or schema changes.

2026.08.26.3Aug 27, 2026

2026.08.26.3

Changed: Retained explicit compilerOptions.strict in deno.json. No behavioral or schema changes.

2026.08.26.1Aug 27, 2026

2026.08.26.1

Changed: Normalized deno.json configuration for repo-wide consistency: added explicit compilerOptions.strict and migrated zod dependency to the import map (bare "zod" specifier instead of inline npm:zod@4.4.3). No behavioral changes — runtime resolution is identical.

2026.08.25.1Aug 26, 2026

2026.08.25.1

Changed: Updated labels for improved extension discoverability. Added cross-cutting category labels (security, observability, finops, infrastructure, networking, compliance, devops, ai, incident-response) where applicable.

updated labels

2026.08.24.2Aug 24, 2026

2026.08.24.2

Added: Output metadata attributes for observability.

durationMs: Method execution duration in milliseconds.
collectedBy: Extension name that produced the data.
fetchedAt: ISO 8601 timestamp when data was fetched (added to resources that previously lacked it).

2026.08.24.1Aug 24, 2026

2026.08.24.1

Fixed source/manifest version mismatch and unquoted manifest version. Added Troubleshooting section documenting record prerequisite, sensitivity array inputs, conditional break-even computation, and independent model state.

2026.08.21.2Aug 22, 2026

2026.08.21.2

Changed: The gpu-capex model's sensitivity method now rejects an empty usefulLifeMonthsRange or empty utilizationPctRange at validation time, with a clear "at least 1 element" error, instead of silently returning a sensitivity matrix with zero rows when either range is passed as [].

No changes to the record, project, or update_hardware_cost methods, and no changes to gpu-cloud, gpu-rental, or the comparison report.

2026.08.21.1Aug 21, 2026

Release Notes

2026.08.21.1

Changed: Added .describe(...) documentation to previously undocumented fields across ScenarioSchema, ProjectionSchema, and (for gpu-capex) SensitivityRowSchema/SensitivitySchema in all three model types (gpu-cloud, gpu-rental, gpu-capex), plus the update_rate, update_hardware_cost, and sensitivity method argument schemas. No behavioral changes.

2026.08.02.1

Fixed: The comparison report declared scope: "workspace", which isn't a valid ReportScope ("method" | "model" | "workflow"). The extension failed to load entirely — swamp doctor extensions reported ValidationFailed for @webframp/cost-projection, and the report could never run.

Changed: The report now runs at model scope and is attached as a default report on gpu-cloud, gpu-rental, and gpu-capex, so it fires after any method call on any instance of the three model types (including sensitivity and update_hardware_cost on gpu-capex, which has no update_rate method). Internally, execute() no longer relies on the single instance's context.dataHandles — it scans every cost-projection instance in the repo via dataRepository.findAllGlobal(), so the table actually compares sibling scenarios (e.g. an on-demand quote against a capacity-block quote) instead of only ever showing the one instance that triggered the run.

Upgrade note: No input/output schema changes. Re-pull the extension to pick up the fix — running any method on a gpu-cloud/gpu-rental/ gpu-capex instance now produces the cross-scenario table, retrievable with swamp report get @webframp/cost-projection-comparison --model <instance>.

2026.08.01.1

Fixed: Report name now uses collective prefix (@webframp/cost-projection-comparison). Previous name (webframp/cost-projection-comparison) failed publish validation.

2026.07.31.1

Initial release.

Three model types: gpu-cloud, gpu-rental, gpu-capex
All normalize to $/GPU-hour for cross-scenario comparison
Capex model includes sensitivity analysis (utilization × useful-life matrix)
Workspace-scoped comparison report with crossover analysis
Single-currency assumption (all scenarios must share a currency)
Optional per-token break-even calculation against API pricing

2026.08.02.1Aug 2, 2026

Release Notes

2026.08.02.1

2026.08.01.1

Fixed: Report name now uses collective prefix (@webframp/cost-projection-comparison). Previous name (webframp/cost-projection-comparison) failed publish validation.

2026.07.31.1

Initial release.

Three model types: gpu-cloud, gpu-rental, gpu-capex
All normalize to $/GPU-hour for cross-scenario comparison
Capex model includes sensitivity analysis (utilization × useful-life matrix)
Workspace-scoped comparison report with crossover analysis
Single-currency assumption (all scenarios must share a currency)
Optional per-token break-even calculation against API pricing

Modified 1 reports

2026.08.01.1Aug 1, 2026

Release Notes

2026.08.01.1

Fixed: Report name now uses collective prefix (@webframp/cost-projection-comparison). Previous name (webframp/cost-projection-comparison) failed publish validation.