Yearly running cost Assuming 12,000 asset requests a year
By hand $585K / year 9,000 hours of work
Subscription n.a.
This workflow $434 – 3.1K / year Machine usage only; setup, hosting and review are extra. Do you want a standalone tool or an asset-library connection? For example, describe a wooden crate, preview its textured draft and export the approved mesh.
One asset request includes up to 2 mesh candidates.
Generate and review asset drafts Generate reviewable drafts from approved references $1,728 – 4,680 setup Connect your asset library Connect approved drafts to an existing workflow $2,304 – 6,360 setup
Which system should it connect to?Blender asset library Game project asset folder An artist approves generated assets before project use. Use authorized references and licensed models. No automatic publishing, production asset overwrites, character rigging or engineering-grade geometry.
What is included Build generation, preview and checked export for simple static props.
Generate static prop drafts, preview them and check their exports. The connected option adds asset-library import, saved versions and an editor connection that leaves originals unchanged.
Reference image scope: 1 approved image per request.
Maximum mesh size: up to 50,000 triangles per candidate.
Maximum texture edge: up to 1,024 pixels.
Reference GPU memory: 24 GB.
Standard independent test set: at least 30 asset requests.
Strict independent test set: at least 60 asset requests.
What standard do you need? Choose the acceptance targets your provider must demonstrate on your own test material.
Standard Test the defined workload and controls. Usable mesh exports: ≥95% Asset checklist satisfied: ≥90% Included Strict Tighter targets and a larger independent test set. Usable mesh exports: ≥99% Asset checklist satisfied: ≥95% Setup +$360 – 1,200 My own requirements Agree different targets and scope with your provider. Quote separately
How these standards are measured Use independent prop briefs and authorized reference images with varied silhouettes, materials and transparency. Include invalid meshes, texture failures, unavailable generators and restricted references. Artists label rubric results independently of prompt tuning. Agree the holdout before tuning. Report each metric with its numerator, denominator and exclusions. Count failed and rejected attempts, not just accepted outputs.
Targets for your selected standard What is checked Target Usable mesh exportsCandidate exports that pass parser and target-import checks divided by all attempted generation requests, including failed requests. ≥95% Asset checklist satisfiedReference checklist items satisfied in independently inspected assets divided by all required checklist items. ≥90% Artist rubric agreementArtist-approved shape and material rubric decisions divided by all independently reviewed rubric decisions. ≥80% Rights and approval controlsPassed reference-rights, export-approval and source-preservation cases divided by all tested control cases; any unauthorized publication fails acceptance. ≥100%
Proposed targets apply to an independently labeled test set on the agreed workload, not every possible future input. Review each pilot result. Any unauthorized action or sensitive-data exposure fails acceptance.
How quickly do you need each asset request? Choose a processing-time target for the agreed reference workload; confirm feasibility with your provider.
Within 10 minutes Included Within 7 minutes Setup +$144 – 480 Within 5 minutes Setup +$288 – 840
Timing details Measure from an accepted brief and reference to checked preview candidates, including queueing, generation, validation and permitted retries. Human art revision and large asset downloads are separate.
The target applies to at least 95% of agreed test runs, with 1 in progress at a time.
How much should each asset request cost? Choose a machine-cost cap for the agreed workload. A lower cap needs testing without relaxing the selected quality targets.
Up to $0.50 Included Up to $0.40 Setup +$144 – 480 Up to $0.30 Setup +$288 – 840
Cost details Includes metered generation compute, bounded prompt assistance, retry allowance, validation and allocated hosting. Paid generation-service minimums, specialist art revision and storage beyond the reference allowance are separate.
Reference machine cost: $0.06 – 0.31 per asset request at 1000 asset requests a month. The selected cap is a target to test, not a replacement for this estimate.
Where do you want it to run? Run the workflow in your cloud or on your own machine. Choose separately whether it may use an external AI model.
Run it onYour cloud Local environment AI model accessApproved model API No external model calls
Data and access details Run in your controlled cloud environment with scoped access and retained review records.
Use an approved external model with the agreed data-processing terms and minimum necessary input.
Local hosting does not itself keep data offline. Approved business-system connections may still use the network. The reference estimate uses licensed image-to-mesh generation on metered compute behind an approved endpoint. Text briefs use an agreed licensed reference-image step within the same compute allowance; generation and failed attempts are metered together. A private model needs license, hardware and quality retesting. Private running cost must be measured in the pilot.
How do you want to use it? Choose a browser interface, an MCP tool for your current agent, or a standalone agent.
Web page Inspect each asset request in a private browser interface. Included Your AI agent Call the workflow through MCP; the same access and approval rules apply. Setup +$72 – 360 Standalone agent Use a dedicated agent with the agreed triggers and controls. Setup +$144 – 480