iilib Workbench
Turn AI pricing and technical specs into decisions.
Workbench tools use the same source-backed model, API and provider data as the iilib Research Engine. Calculations expose their assumptions and the data used.
LLM API Cost Calculator
Estimate per-request and total token cost, compare the same workload across priced models, or start with a dollar budget and estimate how many tokens it can buy.
AI Model Finder
Filter the source-backed model graph by provider, context, API availability, open-weight status, capabilities, modalities and token price.
AI Stack Finder
Start from a tool or agent, separate direct/default from selectable models, follow overlapping API/provider paths, and estimate the same workload cost.
Planned, not yet active
Expand only when the first tool proves useful
Image generation cost
Estimate image-generation spend after iilib has a normalized image-pricing schema.
Open model VRAM
Estimate weight and runtime memory requirements without presenting estimates as benchmarks.
Context window planner
Estimate whether a document and expected output fit inside a model context window.
What Workbench is not
Workbench is not a second pricing database. Every tool should reuse maintained iilib records, show verification freshness, and separate source-backed facts from user-entered assumptions.