> ## Knowledge Base Index
> Fetch the complete knowledge base index at: https://help.100hires.com/sitemap.xml
> Use this file to discover available pages before exploring further.
> Pure-Markdown content can be obtained by appending a '.md' suffix to the content URLs listed in the sitemap (without the trailing slash).

# Pro Model vs Standard: External Research for AI Scoring

# Pro Model vs Standard: External Research for AI Scoring

Understand the difference between standard and premium AI scoring models and when to use each.

## Two AI scoring models

100Hires offers two AI models for candidate scoring:

| Feature | Standard Model | Premium Model |
| ---- | ---- | ---- |
| Analysis depth | Good | 2x smarter |
| Data sources | Internal only | Internal + external research |
| Credit cost | Base rate | ~10x more credits |
| Best for | High volume screening | Senior/critical positions |

## Standard model

The standard model analyzes data available in 100Hires:

**Data sources used:**
* Candidate's resume
* All profile fields
* Questionnaire responses
* Evaluation forms
* Email messages to/from candidate
* Discussion notes
* Job description

**Best for:**
* High-volume initial screening
* Junior to mid-level positions
* When profile data is sufficient
* Conserving AI credits

## Premium model

The premium model does everything the standard model does, plus external research.

**Additional capabilities:**
* Searches external information about candidates
* Researches candidates' past employers
* Finds company details: products, headcount, industry
* Provides deeper context for career history

**What Premium Model Can Find:**
* Company size and growth trajectory of previous employers
* Industry and product focus of past companies
* Whether previous companies were startups, enterprises, etc.
* General company reputation and market position

**Best for:**
* Executive and senior leadership searches
* Roles where company background matters
* When you need context beyond the resume
* Final candidate comparisons

## Enabling premium model

1. Open a job
2. Go to the **AI Scoring** tab
3. Find the premium model toggle
4. Enable: "Premium model (2× smarter; can find external info about candidates' past employers: products, headcount, industry, etc; ~10× AI credits per run)" ![Pro Model vs Standard: External Research for AI Scoring: Enabling premium model](https://storage.crisp.chat/users/helpdesk/website/-/4/f/a/a/4faab8f2b2ab0400/pro-model-vs-standard-ai-scori_f5u0u8.png)
5. Save

## Credit considerations

Premium model uses approximately 10x more credits per candidate scored.

**Example calculation:**
* 100 candidates to score
* Standard model: 100 credits
* Premium model: ~1,000 credits

**Strategy to optimize:**
1. Use standard model for initial screening
2. Reserve premium model for shortlisted candidates
3. Pre-filter with knockout questions before any AI scoring

## When to use each model

### Use standard model when:
* Screening large applicant pools
* Role doesn't require specific company experience
* Budget/credits are limited
* Profile data is comprehensive

### Use premium model when:
* Hiring for leadership positions
* Industry experience matters (need to verify company types)
* Evaluating candidates from unfamiliar companies
* Making final decisions between top candidates
* Company pedigree is a key criterion

## Limitations

Both models can only assess documented information:
* Cannot evaluate speaking ability from resume alone
* Cannot assess personality without interview notes
* Cannot verify claims not documented elsewhere

Premium model adds context but doesn't verify factual accuracy of resumes.

## Related screenshots

These screenshots show the 100Hires areas used in this workflow.

![Candidate Discussion tab zoomed in on the Get AI score button (red outline)](https://100hires.com/wp-content/uploads/2026/08/ai-score-button-zoom-clean.png)

![How to Set Up AI Scoring Criteria and Score Candidates: Setting Up Scoring Criteria](https://storage.crisp.chat/users/helpdesk/website/-/4/f/a/a/4faab8f2b2ab0400/how-to-set-up-ai-scoring-1_6trfe2.png)