Skip to main content

Model Quick Reference

Your at-a-glance guide to choosing the right AI model for every task.

Quick Selection Guide​

Need to decide in 10 seconds?

Your TaskUse This Model
Quick email or translationStandard
Regular business analysisThinking Mini
Complex strategic workThinking – Large
Excel/CSV file analysisData Analysis
PDF or Word document reviewDocument Analysis
Current news or market dataWeb Search (BASIC plan)

Complete Model Comparison Table​

ModelSpeedCostContext WindowBest ForAvailable On
Standard⚡⚡⚡ Fast💰 Low32k tokens (~24k words)Chat, email, translation, quick tasksFREE, LIGHT, BASIC
Thinking Mini⚡⚡ Medium💰💰 Medium64k tokens (~48k words)Daily analysis, reports, reasoningLIGHT, BASIC
Thinking – Large⚡ Slower💰💰💰 High200k tokens (~150k words)Strategic analysis, complex documentsBASIC
Data Analysis⚡⚡ Medium💰💰 Medium64k tokensCSV, Excel, numerical dataLIGHT, BASIC
Document Analysis⚡⚡ Medium💰💰 Medium128k tokens (~96k words)PDFs, Word docs, contractsLIGHT, BASIC
Web Search⚡⚡ Medium💰💰 Medium32k tokensCurrent information, researchBASIC only

Legend:

  • ⚡⚡⚡ = Responses in seconds
  • ⚡⚡ = Balanced speed
  • ⚡ = Takes time but worth it for complex tasks
  • 💰 = Minimal cost per request
  • 💰💰 = Moderate cost
  • 💰💰💰 = Higher cost (use for important tasks)

Model Profiles - One Page​

Standard General Purpose​

When to use:

  • Email drafting and responses
  • Text translation (DE/FR/IT/EN)
  • Simple summarization
  • Quick questions
  • Content formatting
  • Basic writing tasks

Strengths:

  • Very fast response time
  • Low cost per token
  • Good for iteration
  • Excellent multilingual support

Limitations:

  • Less depth in complex analysis
  • Shorter context window
  • Not ideal for long documents

Typical use: "Draft a professional email thanking client for meeting"

Thinking Mini​

When to use:

  • Business memos and reports
  • Meeting summaries
  • Data interpretation
  • Comparative analysis
  • Decision support
  • Regular analytical tasks

Strengths:

  • Balanced reasoning depth and speed
  • Good analytical capabilities
  • Medium context window
  • Cost-effective for daily use

Limitations:

  • Not as deep as Thinking – Large
  • May struggle with very complex multi-step reasoning

Typical use: "Analyze this sales data and identify top 3 trends"

Thinking – Large​

When to use:

  • Strategic planning documents
  • Complex financial analysis
  • Multi-document synthesis
  • High-stakes client deliverables
  • Detailed research
  • Long-form content creation

Strengths:

  • Deepest reasoning capability
  • Largest context window (200k tokens)
  • Handles very complex tasks
  • Excellent for comprehensive analysis

Limitations:

  • Slower response time
  • Higher cost per request
  • Overkill for simple tasks

Typical use: "Create strategic analysis from these 5 reports totaling 100 pages"

Data Analysis​

When to use:

  • Spreadsheet analysis
  • CSV file processing
  • Sales/financial data review
  • Trend identification
  • Statistical summaries
  • Numerical calculations

Strengths:

  • Optimized for structured data
  • Understands tables and formulas
  • Good with numbers
  • Identifies patterns in data

Limitations:

  • Less effective on unstructured text
  • Not designed for creative writing

Typical use: "Analyze this monthly sales CSV and identify products declining >10%"

Document Analysis​

When to use:

  • PDF summarization
  • Contract review
  • Report analysis
  • Multi-document comparison
  • Information extraction
  • Document due diligence

Strengths:

  • Handles long PDFs well
  • Good at extraction
  • Understands document structure
  • Large context window (128k)

Limitations:

  • Not optimized for data/spreadsheets
  • Better for reading than creating

Typical use: "Extract key obligations from this 50-page service contract"

When to use:

  • Market research
  • Current news summaries
  • Recent industry developments
  • Fact-checking current events
  • Competitive intelligence
  • Real-time information needs

Strengths:

  • Access to current web information
  • Can cite recent sources
  • Combines search with reasoning
  • Swiss market focus possible

Limitations:

  • Only available on BASIC plan
  • Results depend on search quality
  • Not a replacement for deep research

Typical use: "Summarize major Swiss fintech developments this month"

Decision Tree​

Start here: What's your primary need?

1. Is your task about current/recent events?​

  • YES → Web Search (BASIC plan required)
  • NO → Continue to #2

2. Are you working with data files (CSV, Excel)?​

  • YES → Data Analysis
  • NO → Continue to #3

3. Are you working with documents (PDF, Word)?​

  • YES → Document Analysis (for reading) or Thinking Large (for analysis)
  • NO → Continue to #4

4. How complex is the reasoning required?​

  • Simple (email, translation, quick question) → Standard
  • Moderate (business memo, data review, analysis) → Thinking Mini
  • Complex (strategy, synthesis, critical decisions) → Thinking – Large

Usage Patterns by Professional Role​

Asset Manager / Wealth Management​

Daily: Standard (emails), Web Search (market news)
Weekly: Thinking Mini (client letters), Data Analysis (portfolio screening)
Monthly: Thinking – Large (investment memos), Document Analysis (research reports)

Retail Business Owner​

Daily: Standard (customer emails), Data Analysis (quick sales check)
Weekly: Data Analysis (inventory), Thinking Mini (marketing content)
Monthly: Thinking Mini (performance reports), Thinking – Large (strategic planning)

Daily: Standard (client emails), Thinking Mini (memos)
Weekly: Document Analysis (contract review), Thinking Large (proposals)
Monthly: Thinking – Large (research synthesis), Web Search (regulatory updates)

Executive / CEO​

Daily: Standard (communications), Web Search (news)
Weekly: Thinking Mini (team updates)
Monthly: Thinking – Large (board materials), Document Analysis (strategic review)

CFO / Finance Team​

Daily: Standard (emails)
Weekly: Data Analysis (variance reports), Thinking Mini (commentary)
Monthly: Thinking – Large (board packs), Data Analysis (budget analysis)

Cost Optimization Strategies​

Start Cheap, Upgrade When Needed​

Workflow pattern:

  1. Draft with Standard or Thinking Mini
  2. Review the output
  3. Refine with same model if good enough
  4. Upgrade to Thinking – Large only if more depth needed

Example:

Round 1 (Standard): "Draft proposal outline" → Fast, cheap outline
Round 2 (Thinking Mini): "Expand section 3" → More detail added
Round 3 (Thinking – Large): "Add strategic analysis" → Deep insights only where needed

Result: 70% cost savings vs. using Thinking – Large for everything

Right-Size Your Model Choice​

Cost comparison for same task:

  • Standard: 1x cost
  • Thinking Mini: 3x cost
  • Thinking – Large: 8x cost

Use expensive models strategically:

  • Client-facing deliverables
  • High-stakes decisions
  • Complex analysis where quality matters most

Use cheaper models freely:

  • Internal communications
  • Rough drafts
  • Exploratory work
  • Iteration and refinement

Batch Similar Tasks​

Efficient batching:

  • Process all emails in one session (Standard)
  • Analyze all data files together (Data Analysis)
  • Review all contracts consecutively (Document Analysis)

Why it saves:

  • Maintains context
  • Reduces switching time
  • Allows pattern reuse

Context Window Guide​

What Fits in Each Model?​

Standard (32k tokens ≈ 24,000 words):

  • ~50 pages of text
  • ~100 emails
  • Short to medium reports

Thinking Mini (64k tokens ≈ 48,000 words):

  • ~100 pages of text
  • Multiple quarterly reports
  • Medium-length contracts

Document Analysis (128k tokens ≈ 96,000 words):

  • ~200 pages of text
  • Long contracts or manuals
  • Multiple related documents

Thinking – Large (200k tokens ≈ 150,000 words):

  • ~300 pages of text
  • Comprehensive research projects
  • Multiple large documents for comparison

Data Analysis (64k tokens):

  • Large spreadsheets (thousands of rows)
  • Multiple CSV files
  • Complex financial models

Practical tip: If your document exceeds the context window, either:

  1. Split it into sections and analyze separately
  2. Upgrade to a model with larger context window
  3. Extract and upload only the relevant portions

Quality Indicators​

When Standard is Enough​

✅ Output is accurate and appropriate
✅ Tone and style are correct
✅ Task is straightforward
✅ Speed matters more than depth

When to Upgrade to Thinking Mini​

⚠️ Standard output is too generic
⚠️ Need more analytical depth
⚠️ Comparative analysis required
⚠️ Multiple factors to consider

When to Use Thinking – Large​

🔴 Thinking Mini lacks strategic insight
🔴 High-stakes client deliverable
🔴 Complex multi-step reasoning needed
🔴 Synthesizing many sources
🔴 Critical business decision

Speed Benchmarks​

Typical response times (approximate):

ModelShort PromptMedium PromptLong Document
Standard2-5 sec5-10 sec10-20 sec
Thinking Mini5-10 sec10-20 sec20-40 sec
Thinking – Large10-20 sec20-40 sec40-90 sec
Data Analysis5-10 sec10-20 sec20-40 sec
Document Analysis5-10 sec10-20 sec20-40 sec
Web Search10-15 sec15-30 sec30-60 sec

Note: Times vary based on prompt complexity, file size, and system load

Plan-Specific Model Access​

FREE Plan​

  • Standard ✅
  • Thinking Mini ❌
  • Thinking – Large ❌
  • Data Analysis ❌
  • Document Analysis ❌
  • Web Search ❌

LIGHT Plan​

  • Standard ✅
  • Thinking Mini ✅
  • Thinking – Large ❌
  • Data Analysis ✅
  • Document Analysis ✅
  • Web Search ❌

BASIC Plan​

  • Standard ✅
  • Thinking Mini ✅
  • Thinking – Large ✅
  • Data Analysis ✅
  • Document Analysis ✅
  • Web Search ✅ (Exclusive to BASIC)

Quick Tips​

Switching Models Mid-Conversation​

You can change models during a conversation!

Example workflow:

  1. Start with Standard to draft outline
  2. Switch to Thinking Mini for section expansion
  3. Switch to Thinking – Large for final strategic polish

The new model sees the full conversation history.

When in Doubt​

Default choices that work for most people:

  • Daily work: Standard + Thinking Mini (LIGHT plan)
  • Professional use: Add Document Analysis + Data Analysis
  • Executive/strategic: Add Thinking – Large + Web Search (BASIC plan)

Testing New Tasks​

When trying something new:

  1. Start with cheaper model
  2. Evaluate quality
  3. Upgrade only if necessary

Printable Quick Reference Card​

╔═══════════════════════════════════════════╗
║     SCHATZI AI MODEL QUICK REFERENCE      ║
╠═══════════════════════════════════════════╣
║                                           ║
║  STANDARD                                 ║
║  • Email, translation, quick tasks        ║
║  • Fast & cheap                           ║
║                                           ║
║  THINKING MINI                            ║
║  • Regular analysis & reports             ║
║  • Daily professional work                ║
║                                           ║
║  THINKING – LARGE                         ║
║  • Strategy & complex analysis            ║
║  • High-stakes deliverables               ║
║                                           ║
║  DATA ANALYSIS                            ║
║  • CSV/Excel files                        ║
║  • Spreadsheet analysis                   ║
║                                           ║
║  DOCUMENT ANALYSIS                        ║
║  • PDF & Word documents                   ║
║  • Contract review                        ║
║                                           ║
║  WEB SEARCH (BASIC only)                  ║
║  • Current news & research                ║
║  • Recent developments                    ║
║                                           ║
╠═══════════════════════════════════════════╣
║  DECISION RULE:                           ║
║  Start simple → upgrade if needed         ║
╚═══════════════════════════════════════════╝
Bookmark This Page

Save this quick reference for fast model selection during your daily work!