Source Filtering

Control which domains, datasets, and sources are included, excluded, or soft-ranked in search results

Control exactly which sources your search uses. Focus on trusted domains, access specific datasets, exclude unreliable sources, or soft-rank results toward preferred domains.

Source filters accept domains, URLs, dataset names, specific paths, or collections.

Save time with Collections! If you use the same source combinations frequently, create a Collection to bundle them and reference by name.

What You Can Do

• Target authoritative sources - Focus on trusted domains and academic datasets
• Block unreliable content - Exclude low-quality or biased sources
• Soft-rank by domain - Boost or demote sources without hard filtering
• Access specific datasets - Search Valyu’s proprietary collections
• Improve result quality - Get more relevant, higher-quality information

Parameters

included_sources

Type: Array of strings

Only search within these sources. Can include domains, URLs, or dataset names.

Example: ["arxiv.org", "valyu/valyu-pubmed"]

excluded_sources

Type: Array of strings

Exclude these sources from results. Same formats as included_sources.

Example: ["example.com", "example.org"]

source_biases

Type: Object (domain → integer)

Soft-rank sources from -5 (strong demotion) to +5 (strong boost) without hard filtering.

Example: {"nasa.gov": 5, "example.com": -3}

If both included_sources and excluded_sources are provided, included_sources takes priority.

When to use source_biases vs hard filters: Use included_sources/excluded_sources when you need strict control over which sources appear. Use source_biases when you want to influence ranking — boosting authoritative domains or demoting low-quality ones — while still allowing all sources to appear if they’re highly relevant.

Source Formats

Format	Example	What It Does
Domain	"arxiv.org"	Includes/excludes entire domain
Base URL	"https://docs.aws.amazon.com"	Includes/excludes entire site
Specific Path	"nasa.gov/news"	Targets only that path
Dataset Name	"valyu/valyu-arxiv"	Searches Valyu’s proprietary datasets
Collection	"collection:my-sources"	Expands to all sources in your saved collection

Path Specificity: When using paths (e.g., "valyu.ai/blog"), only that exact path is affected. For entire domains, use just the domain name.

Examples

Academic Sources

Python

from valyu import Valyu

valyu = Valyu("your-api-key-here")
response = valyu.search(
    "quantum computing error correction",
    included_sources=[
        "valyu/valyu-arxiv",
        "valyu/valyu-pubmed",
        "valyu/valyu-biorxiv",
        "valyu/valyu-medrxiv"
    ],
    search_type="all"
)

Exclude Non-Government Sources

Python

from valyu import Valyu

valyu = Valyu("your-api-key-here")
response = valyu.search(
    "artificial intelligence safety research",
    excluded_sources=[
        "example.com",
        "example.org",
        "example.net"
    ]
)

Official Documentation

Python

from valyu import Valyu

valyu = Valyu("your-api-key-here")
response = valyu.search(
    "React server components best practices",
    included_sources=[
        "https://react.dev/",
        "https://nextjs.org/docs",
        "https://docs.aws.amazon.com/",
        "developer.mozilla.org"
    ]
)

Use Cases

Financial Research

Python

from valyu import Valyu

valyu = Valyu("your-api-key-here")
response = valyu.search(
    "cryptocurrency regulation impact banking sector",
    included_sources=[
        "federalreserve.gov",
        "sec.gov", 
        "sec.gov/cgi-bin/browse-edgar",
        "treasury.gov",
        "imf.org"
    ],
    max_num_results=15
)

Medical Research

Python

from valyu import Valyu

valyu = Valyu("your-api-key-here")
response = valyu.search(
    "immunotherapy cancer treatment efficacy",
    included_sources=[
        "valyu/valyu-pubmed",
        "valyu/valyu-clinical-trials",
        "valyu/valyu-drug-labels",
        "valyu/valyu-medrxiv"
    ],
    search_type="proprietary"
)

Technical Documentation

Python

from valyu import Valyu

valyu = Valyu("your-api-key-here")
response = valyu.search(
    "Kubernetes security best practices RBAC",
    included_sources=[
        "kubernetes.io/docs",
        "docs.aws.amazon.com",
        "cloud.google.com/kubernetes-engine/docs"
    ]
)

News

Python

from valyu import Valyu

valyu = Valyu("your-api-key-here")
response = valyu.search(
    "artificial intelligence regulation European Union",
    included_sources=[
        "scholar.google.com",
        "gov.uk/search/news",
        "who.int",
        "europa.eu",
        "politico.eu"
    ]
)

Source Biases

Source biases let you influence ranking without hard filtering. Unlike included_sources/excluded_sources, biased sources can still appear (or be absent) based on relevance — biases just nudge the ranking.

Values range from -5 (strong demotion) to +5 (strong boost). A value of 0 has no effect.

Boost Government Sources

Python

from valyu import Valyu

valyu = Valyu("your-api-key-here")
response = valyu.search(
    "climate change policy impact",
    source_biases={
        "epa.gov": 5,
        "nasa.gov": 4,
        "noaa.gov": 3,
        "nih.gov": 2,
        "example.com": -4
    }
)

Combine with Hard Filters

You can use source_biases alongside included_sources or excluded_sources for fine-grained control:

Python

from valyu import Valyu

valyu = Valyu("your-api-key-here")
response = valyu.search(
    "federal research funding allocation",
    excluded_sources=["example.com", "example.org"],
    source_biases={
        "nsf.gov": 5,
        "nih.gov": 4,
        "energy.gov": 3
    }
)