List BoaviztAPI's impact criteria with their code, unit and meaning, e.g. gwp (climate change, kgCO2eq), adp (abiotic resources, kgSbeq) or pe (primary energy, MJ).

Use it to explain a code that appears in a result, or to choose criteria for the estimate tools; it estimates nothing. query filters on code, unit or description words such as 'water'. Read-only; the list comes from BoaviztAPI and is cached for a day.

Find the exact names BoaviztAPI knows, before estimating.

Use it whenever an estimate needs a name you are unsure of: a GPU or CPU model, a cloud instance type (needs provider), a cloud region, a server or device archetype (needs device_type), or a country. Results are exact upstream strings, including upstream typos such as 'platfom_gpucompute_high', best matches first and capped by limit; total_matches says how many exist. To see what one item contains, use describe_catalog_item instead. Read-only; lists are cached for a day.

Show what a named catalog item contains: a cloud instance's vCPUs, memory and host platform; a server or device archetype's default parts and usage; or a GPU's or CPU's specifications.

Use it to explain or check the assumptions behind an estimate, or to choose between similar items; it does not estimate impacts. name is resolved to the exact upstream name, and an ambiguous name returns the candidates (use search_catalog to find names). provider is required for cloud_instance and device_type for device_archetype. Read-only; cached for a day.

Estimate the environmental footprint of one cloud VM: its share of the host server's manufacturing (embedded) and the electricity it uses (use), per impact criterion, with units, min-max ranges and warnings.

Use it for a single instance; use compare_estimates to rank several and estimate_server for physical servers. Give either region (the provider's region, which fixes the electricity mix) or usage_location (a country), not both; with neither, a wide European default is used. duration_hours sets the period (8760 = one year); omit it for the whole hardware lifetime. Read-only: one BoaviztAPI request, cached for an hour; if one criterion fails upstream the others are still returned. Use search_catalog(kind='cloud_instance', provider=...) if unsure of a name.

Estimate the environmental footprint of an on-premise server, optionally with GPUs: manufacturing (embedded) and electricity (use) impacts per criterion, with units, ranges and warnings.

Use it for physical servers; use estimate_cloud_instance for cloud VMs and estimate_component for one part on its own. Any part you leave out (CPU, RAM, disks, GPU, power supply, case) comes from archetype (default 'platform_compute_medium'), and the result lists which ones. The default lifetime is 4 years (35040 h); a duration_hours longer than the lifetime counts manufacturing once and says so. Set usage_location for a country's electricity mix and detail='breakdown' for per-component impacts. Read-only: one BoaviztAPI request, cached for an hour.

Estimate the environmental footprint of one kind of component (gpu, cpu, ram, ssd or hdd), units of it: manufacturing (embedded) and electricity (use) impacts per criterion, with units, ranges and warnings.

Use it for parts on their own, such as eight H100 GPUs. A GPU alone excludes its host server: for a whole machine use estimate_server with a gpu block. name and tdp_w apply to gpu and cpu, capacity_gb to ram, ssd and hdd; other combinations are an error. Names are resolved to BoaviztAPI's model names and an ambiguous one (such as 'h100') returns the candidates. Read-only: one BoaviztAPI request, cached for an hour.

Estimate the environmental footprint of an end-user device: laptop, desktop, smartphone, tablet, television, box, vr_headset, vr_controller, monitor, usb_stick, external_ssd, external_hdd or iot_device. Returns manufacturing (embedded) and electricity (use) impacts per criterion, with units, ranges and warnings.

Use it for devices people use, not servers or cloud (see estimate_server and estimate_cloud_instance). Without archetype the device type's default profile is used, except for iot_device, which needs one; list them with search_catalog(kind='device_archetype', device_type=...). use_time_ratio is the share of time the device is on (0 to 1). Criteria without data are reported as not available, never as 0. Read-only: one BoaviztAPI request, cached for an hour.

Compare 2 to 5 options (cloud instances, servers, components or devices) over one common period and rank them per criterion, with each total, the difference from the first option, and whether the uncertainty ranges overlap.

Use it whenever the question is which option is greener or how options differ; use the single estimate tools for one item. Each item has a label, a kind, and params holding the matching estimate tool's inputs (without duration_hours, criteria or detail). duration_hours is required so items with different lifetimes compare on equal terms. Every item is checked before any estimate runs, and an invalid item fails the call with its label named. Read-only: one BoaviztAPI request per item.