PubChem Compound Research — tool reference

Server URL: https://toolcargo.com/mcp/pubchem · Scope: connector:pubchem. Activate Site Audit free and connect with ToolCargo OAuth or a scoped API key. No PubChem account or provider key. Each successful operation uses one shared-plan call.

pubchem_resolve_name

{ "name": "aspirin", "limit": 5 }

Supply one complete chemical name, up to 160 characters. PubChem resolves its indexed names and synonyms; a result is a compound candidate, not proof of the chemistry you intended. Multiple candidates remain explicit. Inspect an exact CID before using its properties. Names are sent to PubChem, so use public research terms.

The tool returns up to ten candidates from a bounded source list, with the full valid candidate count, ambiguity and truncation. A source list exceeding the supported bound requires a more specific name. There is no invented pagination, similarity search, structure input, listkey polling or automatic candidate selection. A recognized name-match absence describes this name lookup, not chemical nonexistence.

pubchem_compound_properties

{ "cid": "2244" }

Use an exact positive decimal CID string without leading zeros. ToolCargo first requests identifying CID/title metadata, then the selected property table. Returned identities must match. These separate requests are not an atomic snapshot. Contributed descriptions and their links are discarded.

If PubChem supplies no identifying description metadata, output reports identity_unavailable with existence unknown. It does not infer an absent compound from that response or from an empty property table. Missing individual properties remain missing. Identity title and property title are kept separate.

Read the properties in context

Selected fields are the source summary title, molecular formula, molecular weight, InChIKey and IUPAC name. Molecular weight is a calculated value in g/mol, with original decimal text or source numeric type retained. It is not a measured sample mass, purity, monoisotopic mass or dose. Names and identifiers do not establish equivalence between different stereochemical records.

No structures, images, assays, supplier listings, safety annotations, synthesis instructions or medical recommendations. Retrieval times describe ToolCargo’s source requests. Source record-update time, immutable revision, property-method version and record-specific license are unavailable and stay unknown.

Source access and request bounds

Fixed PubChem PUG REST HTTPS endpoints only, with no account identifiers or credentials forwarded. Shared conservative pacing and source throttle feedback apply across ToolCargo instances. Calls can return a busy response while a source cooldown is active; repeat explicitly later. No automatic retries, redirects, result cache, stale fallback or followed links. Private source responses are not stored in a shared result cache.

The source’s request-count and server-runtime ceilings are upper limits, not guaranteed capacity. ToolCargo checks all three source throttle indicators and retains the strongest pause and Retry-After instruction. Bounded responses and operation deadlines can stop a lookup rather than return an unchecked record.

Attribution and reuse

Source: PubChem, National Center for Biotechnology Information, U.S. National Library of Medicine. ToolCargo selects, validates and formats source metadata; its output is not NCBI-authored material or endorsement. PubChem source access is free. ToolCargo fees concern hosted access and processing, while contributor rights and source notices remain applicable. No blanket CC0 or unrestricted reuse permission is claimed.

Primary references: PUG REST · Dynamic throttling · Source downloads and rights · NCBI policies and disclaimer.