NeuralScio searches across 10+ scholarly sources. The registered adapters have different access profiles. They expose different combinations of metadata, links, open PDFs, citation context, and policy constraints.
Implemented Providers
| Provider | Primary use | Access profile | | --- | --- | --- | | arXiv | Open preprints in physics, math, CS, AI | Open access | | bioRxiv | Biology preprints | Open access | | medRxiv | Medical preprints | Open access | | PubMed | Biomedical literature | Metadata plus links | | Semantic Scholar | Broad discovery and citation metadata | Metadata-focused; optional API key improves access/rate limits | | OpenAlex | Broad scholarly catalog | Metadata plus links | | Crossref | DOI and metadata lookup | Metadata-focused | | Europe PMC | Biomedical literature with open-access coverage | Metadata plus OA links | | IEEE Xplore | Engineering and CS literature | Metadata-focused, API/subscription dependent, often restricted | | ACM Digital Library | Computing literature | Metadata-focused, access dependent, often restricted | | ChinaRxiv | Chinese preprints with English-access metadata and translated full text where available | Metadata plus full text/PDF links; some records may be machine-translated | | SovietRxiv | Translated Soviet-era research papers | Metadata plus full text/PDF links; historical source dates may be preserved |
Why Provider Mix Matters
Open-access sources
arXiv, bioRxiv, and medRxiv are valuable when you want faster hand-off into PDF reading and AI work because full text is more often directly reachable.
Metadata-heavy sources
PubMed, Semantic Scholar, OpenAlex, Crossref, IEEE, and ACM are still useful, but the workflow may move through metadata, links, and policy limits rather than immediate full text.
Provider Policy Reality
Not every source exposes the same AI or offline path.
In practice:
- some providers are best for discovery and metadata
- some are better for open full text
- some have policy or access restrictions that narrow downstream AI or caching use
- Semantic Scholar and IEEE can use provider API keys when configured
- PubMed can use an NCBI API key for better request pacing
- OpenAlex and Crossref support polite metadata access and can include contact email settings
- provider policies can prevent automatic download, full-text handling, or AI processing for restricted sources
- ChinaRxiv exposes public API access without an API key; some records are machine-translated and should be treated accordingly
- SovietRxiv exposes public API access where available; historical source dates and translated text should be interpreted accordingly
Practical Guidance
1. Use several providers when coverage matters. 2. Prefer open-access-heavy sources when you need fast PDF and AI continuity. 3. Treat IEEE and ACM as important discovery surfaces, not guaranteed full-text pipelines. 4. Use Crossref when DOI lookup matters. 5. Use OpenAlex and Semantic Scholar when you want broad discovery plus citation context. 6. Do not assume every provider result can be downloaded, cached, or sent to AI.