Which AI is better for coding in 2026?

Claude is definitively better for coding. Claude Opus 4.5 achieved 80.9% accuracy on SWE-bench Verified, significantly outperforming ChatGPT's ~70%. Blind tests consistently show Claude winning on coding tasks and reasoning-heavy work. However, ChatGPT excels in speed and quick generation tasks.

Does Claude have a larger context window than ChatGPT?

Yes. Claude Pro offers a 200K token context window, compared to ChatGPT's standard 128K tokens at the paid tier. This 56% advantage makes Claude superior for handling long documents, large codebases, and extended analysis tasks.

Can Claude generate images like ChatGPT?

No. As of 2026, Claude does not have native image generation capabilities. ChatGPT integrates DALL-E 3 for image creation. However, Claude added web search in February 2026, closing a previous gap. If you need image generation, ChatGPT is the clear choice.

Which AI is better for creative writing?

ChatGPT is better for creative writing and content generation. It performs exceptionally well in blind tests for creative tasks and has better memory retention for maintaining narrative consistency. Its broader feature set and ecosystem also support multimedia content creation.

Should I use both ChatGPT and Claude?

Yes, using both is recommended if budget allows. Many developers use both in tools like Playcode, which aggregates 15+ AI models in one interface. Use Claude for coding, analysis, and long documents; use ChatGPT for creative work, image generation, and quick tasks. This hybrid approach leverages each tool's strengths.

ChatGPT vs Claude

Updated May 22, 2026

ChatGPT

OpenAI's advanced AI assistant available in free and premium tiers with multimodal capabilities.

Content creators, designers, marketers, and users needing image generation or creative tasks

Check Price

Claude

Anthropic's advanced AI assistant optimized for coding, reasoning, and nuanced content generation.

Software engineers, data analysts, researchers, and professionals handling large documents or complex coding tasks

Short Answer

Claude excels in coding and deep analysis with superior benchmarks (80.9% vs ~70% on SWE-bench), while ChatGPT leads in multimodal features like image generation and has broader ecosystem integration. The choice depends on your primary use case: choose Claude for software engineering and reasoning tasks, ChatGPT for creative and visual work.

Our Verdict

Both ChatGPT and Claude are leading AI assistants in 2026, each dominating different niches. Claude is the superior choice for developers and technical analysis, while ChatGPT's broader capabilities and creative features make it ideal for content creation and multimedia tasks. Neither is universally better—success depends on aligning the tool with your specific needs.

ChatGPT6.7

8.3Claude

Choose ChatGPT if

Content creators, designers, marketers, and users needing image generation or creative tasks

Track this comparison

Get notified when prices change, new specs ship, or our verdict updates.

Triggers: price change new spec verdict update

No spam. Stop anytime.

Key Differences at a Glance

⚡

Coding Performance: Claude wins (80.9% (SWE-bench Verified) vs ~70% (SWE-bench Verified))

🧠

Context Window (Paid Tier): Claude wins (200K tokens vs 128K tokens)

📅

Native Image Generation: ChatGPT wins (Yes (DALL-E integration) vs No (added web search in 2026))

See all 7 differences

Key Facts & Figures

Metric	ChatGPT	Claude	Diff
Pro Tier Monthly Cost(USD)	$20	—	—
Plus Subscription Price(USD/month)	$20	—	—
SWE-bench Verified Score(%)	~70%	80.9%	-13%
Context Window (Paid Tier)(tokens)	128,000

Key Differences

ChatGPT

Attribute

Claude

~70% (SWE-bench Verified)

Coding Performance

80.9% (SWE-bench Verified)🏆

128K tokens

Context Window (Paid Tier)

200K tokens🏆

Yes (DALL-E integration)🏆

Related Comparisons

Perplexity vs ChatGPT

software

Microsoft Copilot vs ChatGPT

software

DeepSeek vs ChatGPT

software

Claude vs Gemini

technology

Grok vs ChatGPT

technology

Microsoft Copilot vs ChatGPT

software

ChatGPT vs Meta AI

technology

WordPress vs Wix

software

Canva vs Photoshop

software

Slack vs Microsoft Teams

software

Midjourney vs DALL-E

software

Figma vs Sketch

software

technology

Best AI Assistants 2026: ChatGPT vs Claude vs Gemini vs Perplexity

Discover the top AI assistants of 2026 and find the perfect match for your needs. We compare ChatGPT, Claude, Gemini, and Perplexity across features, pricing, and performance to help you choose the best AI companion.

technology

Best AI Chatbots 2026: ChatGPT vs Claude vs Gemini vs Perplexity

Discover the top AI chatbots of 2026 and how they compare. We analyze ChatGPT, Claude, Gemini, and Perplexity across features, pricing, and performance to help you choose the best AI assistant for your needs.

Explore Entities

More Software

People Also Compare

Last updated: May 22, 2026AI generated

Coding Performance

ChatGPT

~70% (SWE-bench Verified)

Claude

80.9% (SWE-bench Verified)🏆

Context Window (Paid Tier)

ChatGPT

128K tokens

Claude

200K tokens🏆

Native Image Generation

ChatGPT

Yes (DALL-E integration)🏆

Claude

No (added web search in 2026)

Web Search Capability

ChatGPT

Yes

Claude

Yes (added February 2026)

Long Document Analysis

ChatGPT

Moderate

Claude

Superior🏆

Multimodal Features

ChatGPT

Extensive (image gen, vision, audio)🏆

Claude

Limited (vision only)

Speed & Agentic Workflows

ChatGPT

Faster execution🏆

Claude

Slower but more accurate

Full Comparison

Attribute	ChatGPT	Claude

Monthly Active Users(millions)	200M+	Growing

Real-Time Web Search(availability)	Paid ($20+/mo)	—
Web Search	Yes	Yes (Feb 2026)
Free Tier Web Access	Limited or subscription-only	—
Image Generation	Native (DALL-E 3)	Not available
Real-Time Web Access (Free Tier)	No	—
Show 7 more attributes Screen/Visual Analysis Yes, via GPT-4 Vision in Plus/Pro tiers — Multimodal Support Text, image, voice, video — Real-Time Data Access No (knowledge cutoff) — Multimodal Capability Score(capability level) Advanced (image, video, text, document analysis) — Platform Integration Points(platforms) 1 (standalone web/app) — Native Microsoft 365 Integration Manual copy-paste only — Multimodal Capabilities Text and code focused —

Free Tier Quality	GPT-3.5 (basic)	—

Coding Blind Test Winner	Not favored	Won coding tests

Long Document Handling	Moderate	Superior
Context Window (Paid Tier)(tokens)	128,000	200,000
Reasoning Task Performance	Very Strong (GPT-4o)	—
Complex Reasoning Accuracy(performance rating)	GPT-4o: Superior accuracy on technical/logic problems	—
Creative Writing Instruction Adherence(quality rating)	Excellent - follows detailed creative prompts precisely	—

Execution Speed	Faster	Slower (more thorough)
Reasoning Benchmark Score(percentile)	Strong (GPT-4o/5.4 capable)	—
M365 Task Accuracy Rate(%)	82%	—

Market Adoption 2026	Broader ecosystem	Growing adoption

Primary Use Case	Conversational assistance & content generation	—
Research Organization Features	Projects feature	—

Reasoning Model Capability	o1/o3 advanced reasoning series	—

Pro Tier Monthly Cost(USD)	$20	—
Plus Subscription Price(USD/month)	$20	—
Minimum Monthly Cost for Advanced Features(USD)	$20 (Plus tier)	—
API Cost (Input Tokens)($ per million tokens)	$0.05-3.00 (varies by model)	—
Minimum Monthly Cost (Consumer)($)	$5 (ChatGPT Go)	—
Show 2 more attributes Subscription Cost (Annual)(USD) $240 (ChatGPT Plus) — Premium Monthly Cost(USD) $20/month —

Third-Party Integration Ecosystem	Hundreds of plugins and integrations	—

Real-Time Information Access	Knowledge cutoff with occasional browsing	—
Real-Time Information Advantage(hours)	Web access in Plus tier	—

Source Citation System	Available but not primary focus	—

Model Customization Options	Custom GPTs with established ecosystem	—

Parallel Model Processing	Single model per session	—

Memory & Context Window(tokens)	GPT-5.2 advanced context	—

Information Synthesis Quality(rating)	Superior synthesis and judgment	—
Content Writing Refinement	Nuanced and contextual	—

SWE-bench Verified Score(%)	~70%	80.9%

Base AI Model	GPT-4, GPT-4o, o1, and newer models	—
Base Model Version	GPT-5.2	—

Context Window Size(tokens)	Up to 200,000 tokens (o1 models)	—
Context Window(tokens)	200K+ tokens	200K+

Custom GPT/Agent Creation	Yes; extensive custom GPT ecosystem	—

Third-Party App Integrations(count)	Thousands via Zapier and API	—
Ecosystem Integration	Standalone platform	—

Enterprise SLA & Compliance	Yes (ChatGPT Enterprise custom terms)	—

On-Premise Deployment	Limited (enterprise only)	—

Third-Party Integrations(score/10)	Extensive (1000+ integrations)	—

User Interface Rating(category)	Consumer-friendly, polished	—

Monthly Active Users(millions)	200+	Claude

Monthly Website Visits(millions)	1,500+	—

Launch Date	November 2022	—

Platform Availability	Web, iOS, Android, API	—
Free Tier Availability	Free basic features	—

Primary User Age Demographic(% under 35)	Mixed	—

Free Tier Message Limit(messages/month)	40 messages/month (ChatGPT Go)	—
Monthly Premium Cost(USD)	$20 (Plus) / $200 (Pro)	—
Cost for Annual Pro Usage(USD/year)	$240 (Plus annually)	—

Setup Time(days)	5-10 (account creation, app download)	—

Third-Party App Connections(apps)	Thousands via Zapier	—

Windows OS Integration	Separate application	—

Data Privacy Model	Processes public web data only	—

Developer Preference Rating(%)	Claude	—

Coding Task Excellence(benchmark)	Excellent (leading tool)	—
Reasoning Quality(benchmark)	Superior (Claude 4.6)	—

Company Valuation(billions USD)	Claude	—

ChatGPT

Claude

Monthly Active Users(millions)

200M+

Growing

Real-Time Web Search(availability)

Paid ($20+/mo)

—

Web Search

Yes

Yes (Feb 2026)

Free Tier Web Access

Limited or subscription-only

—

Image Generation

Native (DALL-E 3)

Not available

Real-Time Web Access (Free Tier)

—

Show 7 more attributes

Screen/Visual Analysis

Yes, via GPT-4 Vision in Plus/Pro tiers

—

Multimodal Support

Text, image, voice, video

—

Real-Time Data Access

No (knowledge cutoff)

—

Multimodal Capability Score(capability level)

Advanced (image, video, text, document analysis)

—

Platform Integration Points(platforms)

1 (standalone web/app)

—

Native Microsoft 365 Integration

Manual copy-paste only

—

Multimodal Capabilities

Text and code focused

—

Free Tier Quality

GPT-3.5 (basic)

—

Coding Blind Test Winner

Not favored

Won coding tests

Long Document Handling

Moderate

Superior

Context Window (Paid Tier)(tokens)

128,000

200,000

Reasoning Task Performance

Very Strong (GPT-4o)

—

Complex Reasoning Accuracy(performance rating)

GPT-4o: Superior accuracy on technical/logic problems

—

Creative Writing Instruction Adherence(quality rating)

Excellent - follows detailed creative prompts precisely

—

Execution Speed

Faster

Slower (more thorough)

Reasoning Benchmark Score(percentile)

Strong (GPT-4o/5.4 capable)

—

M365 Task Accuracy Rate(%)

82%

—

Market Adoption 2026

Broader ecosystem

Growing adoption

Primary Use Case

Conversational assistance & content generation

—

Research Organization Features

Projects feature

—

Reasoning Model Capability

o1/o3 advanced reasoning series

—

Pro Tier Monthly Cost(USD)

$20

—

Plus Subscription Price(USD/month)

$20

—

Minimum Monthly Cost for Advanced Features(USD)

$20 (Plus tier)

—

API Cost (Input Tokens)($ per million tokens)

$0.05-3.00 (varies by model)

—

Minimum Monthly Cost (Consumer)($)

$5 (ChatGPT Go)

—

Show 2 more attributes

Subscription Cost (Annual)(USD)

$240 (ChatGPT Plus)

—

Premium Monthly Cost(USD)

$20/month

—

Third-Party Integration Ecosystem

Hundreds of plugins and integrations

—

Real-Time Information Access

Knowledge cutoff with occasional browsing

—

Real-Time Information Advantage(hours)

Web access in Plus tier

—

Source Citation System

Available but not primary focus

—

Model Customization Options

Custom GPTs with established ecosystem

—

Parallel Model Processing

Single model per session

—

Memory & Context Window(tokens)

GPT-5.2 advanced context

—

Information Synthesis Quality(rating)

Superior synthesis and judgment

—

Content Writing Refinement

Nuanced and contextual

—

SWE-bench Verified Score(%)

~70%

80.9%

Base AI Model

GPT-4, GPT-4o, o1, and newer models

—

Base Model Version

GPT-5.2

—

Context Window Size(tokens)

Up to 200,000 tokens (o1 models)

—

Context Window(tokens)

200K+ tokens

200K+

Custom GPT/Agent Creation

Yes; extensive custom GPT ecosystem

—

Third-Party App Integrations(count)

Thousands via Zapier and API

—

Ecosystem Integration

Standalone platform

—

Enterprise SLA & Compliance

Yes (ChatGPT Enterprise custom terms)

—

On-Premise Deployment

Limited (enterprise only)

—

Third-Party Integrations(score/10)

Extensive (1000+ integrations)

—

User Interface Rating(category)

Consumer-friendly, polished

—

Monthly Active Users(millions)

200+

Claude

Monthly Website Visits(millions)

1,500+

—

Launch Date

November 2022

—

Platform Availability

Web, iOS, Android, API

—

Free Tier Availability

Free basic features

—

Primary User Age Demographic(% under 35)

Mixed

—

Free Tier Message Limit(messages/month)

40 messages/month (ChatGPT Go)

—

Monthly Premium Cost(USD)

$20 (Plus) / $200 (Pro)

—

Cost for Annual Pro Usage(USD/year)

$240 (Plus annually)

—

Setup Time(days)

5-10 (account creation, app download)

—

Third-Party App Connections(apps)

Thousands via Zapier

—

Windows OS Integration

Separate application

—

Data Privacy Model

Processes public web data only

—

Developer Preference Rating(%)

Claude

—

Coding Task Excellence(benchmark)

Excellent (leading tool)

—

Reasoning Quality(benchmark)

Superior (Claude 4.6)

—

Company Valuation(billions USD)

Claude

—

Full Comparison

Attribute	ChatGPT	Claude

Monthly Active Users(millions)	200M+	Growing

Real-Time Web Search(availability)	Paid ($20+/mo)	—
Web Search	Yes	Yes (Feb 2026)
Free Tier Web Access	Limited or subscription-only	—
Image Generation	Native (DALL-E 3)	Not available
Real-Time Web Access (Free Tier)	No	—
Show 7 more attributes Screen/Visual Analysis Yes, via GPT-4 Vision in Plus/Pro tiers — Multimodal Support Text, image, voice, video — Real-Time Data Access No (knowledge cutoff) — Multimodal Capability Score(capability level) Advanced (image, video, text, document analysis) — Platform Integration Points(platforms) 1 (standalone web/app) — Native Microsoft 365 Integration Manual copy-paste only — Multimodal Capabilities Text and code focused —

Free Tier Quality	GPT-3.5 (basic)	—

Coding Blind Test Winner	Not favored	Won coding tests

Long Document Handling	Moderate	Superior
Context Window (Paid Tier)(tokens)	128,000	200,000
Reasoning Task Performance	Very Strong (GPT-4o)	—
Complex Reasoning Accuracy(performance rating)	GPT-4o: Superior accuracy on technical/logic problems	—
Creative Writing Instruction Adherence(quality rating)	Excellent - follows detailed creative prompts precisely	—

Execution Speed	Faster	Slower (more thorough)
Reasoning Benchmark Score(percentile)	Strong (GPT-4o/5.4 capable)	—
M365 Task Accuracy Rate(%)	82%	—

Market Adoption 2026	Broader ecosystem	Growing adoption

Primary Use Case	Conversational assistance & content generation	—
Research Organization Features	Projects feature	—

Reasoning Model Capability	o1/o3 advanced reasoning series	—

Pro Tier Monthly Cost(USD)	$20	—
Plus Subscription Price(USD/month)	$20	—
Minimum Monthly Cost for Advanced Features(USD)	$20 (Plus tier)	—
API Cost (Input Tokens)($ per million tokens)	$0.05-3.00 (varies by model)	—
Minimum Monthly Cost (Consumer)($)	$5 (ChatGPT Go)	—
Show 2 more attributes Subscription Cost (Annual)(USD) $240 (ChatGPT Plus) — Premium Monthly Cost(USD) $20/month —

Third-Party Integration Ecosystem	Hundreds of plugins and integrations	—

Real-Time Information Access	Knowledge cutoff with occasional browsing	—
Real-Time Information Advantage(hours)	Web access in Plus tier	—

Source Citation System	Available but not primary focus	—

Model Customization Options	Custom GPTs with established ecosystem	—

Parallel Model Processing	Single model per session	—

Memory & Context Window(tokens)	GPT-5.2 advanced context	—

Information Synthesis Quality(rating)	Superior synthesis and judgment	—
Content Writing Refinement	Nuanced and contextual	—

SWE-bench Verified Score(%)	~70%	80.9%

Base AI Model	GPT-4, GPT-4o, o1, and newer models	—
Base Model Version	GPT-5.2	—

Context Window Size(tokens)	Up to 200,000 tokens (o1 models)	—
Context Window(tokens)	200K+ tokens	200K+

Custom GPT/Agent Creation	Yes; extensive custom GPT ecosystem	—

Third-Party App Integrations(count)	Thousands via Zapier and API	—
Ecosystem Integration	Standalone platform	—

Enterprise SLA & Compliance	Yes (ChatGPT Enterprise custom terms)	—

On-Premise Deployment	Limited (enterprise only)	—

Third-Party Integrations(score/10)	Extensive (1000+ integrations)	—

User Interface Rating(category)	Consumer-friendly, polished	—

Monthly Active Users(millions)	200+	Claude

Monthly Website Visits(millions)	1,500+	—

Launch Date	November 2022	—

Platform Availability	Web, iOS, Android, API	—
Free Tier Availability	Free basic features	—

Primary User Age Demographic(% under 35)	Mixed	—

Free Tier Message Limit(messages/month)	40 messages/month (ChatGPT Go)	—
Monthly Premium Cost(USD)	$20 (Plus) / $200 (Pro)	—
Cost for Annual Pro Usage(USD/year)	$240 (Plus annually)	—

Setup Time(days)	5-10 (account creation, app download)	—

Third-Party App Connections(apps)	Thousands via Zapier	—

Windows OS Integration	Separate application	—

Data Privacy Model	Processes public web data only	—

Developer Preference Rating(%)	Claude	—

Coding Task Excellence(benchmark)	Excellent (leading tool)	—
Reasoning Quality(benchmark)	Superior (Claude 4.6)	—

Company Valuation(billions USD)	Claude	—

ChatGPT

Claude

Monthly Active Users(millions)

200M+

Growing

Real-Time Web Search(availability)

Paid ($20+/mo)

—

Web Search

Yes

Yes (Feb 2026)

Free Tier Web Access

Limited or subscription-only

—

Image Generation

Native (DALL-E 3)

Not available

Real-Time Web Access (Free Tier)

—

Show 7 more attributes

Screen/Visual Analysis

Yes, via GPT-4 Vision in Plus/Pro tiers

—

Multimodal Support

Text, image, voice, video

—

Real-Time Data Access

No (knowledge cutoff)

—

Multimodal Capability Score(capability level)

Advanced (image, video, text, document analysis)

—

Platform Integration Points(platforms)

1 (standalone web/app)

—

Native Microsoft 365 Integration

Manual copy-paste only

—

Multimodal Capabilities

Text and code focused

—

Free Tier Quality

GPT-3.5 (basic)

—

Coding Blind Test Winner

Not favored

Won coding tests

Long Document Handling

Moderate

Superior

Context Window (Paid Tier)(tokens)

128,000

200,000

Reasoning Task Performance

Very Strong (GPT-4o)

—

Complex Reasoning Accuracy(performance rating)

GPT-4o: Superior accuracy on technical/logic problems

—

Creative Writing Instruction Adherence(quality rating)

Excellent - follows detailed creative prompts precisely

—

Execution Speed

Faster

Slower (more thorough)

Reasoning Benchmark Score(percentile)

Strong (GPT-4o/5.4 capable)

—

M365 Task Accuracy Rate(%)

82%

—

Market Adoption 2026

Broader ecosystem

Growing adoption

Primary Use Case

Conversational assistance & content generation

—

Research Organization Features

Projects feature

—

Reasoning Model Capability

o1/o3 advanced reasoning series

—

Pro Tier Monthly Cost(USD)

$20

—

Plus Subscription Price(USD/month)

$20

—

Minimum Monthly Cost for Advanced Features(USD)

$20 (Plus tier)

—

API Cost (Input Tokens)($ per million tokens)

$0.05-3.00 (varies by model)

—

Minimum Monthly Cost (Consumer)($)

$5 (ChatGPT Go)

—

Show 2 more attributes

Subscription Cost (Annual)(USD)

$240 (ChatGPT Plus)

—

Premium Monthly Cost(USD)

$20/month

—

Third-Party Integration Ecosystem

Hundreds of plugins and integrations

—

Real-Time Information Access

Knowledge cutoff with occasional browsing

—

Real-Time Information Advantage(hours)

Web access in Plus tier

—

Source Citation System

Available but not primary focus

—

Model Customization Options

Custom GPTs with established ecosystem

—

Parallel Model Processing

Single model per session

—

Memory & Context Window(tokens)

GPT-5.2 advanced context

—

Information Synthesis Quality(rating)

Superior synthesis and judgment

—

Content Writing Refinement

Nuanced and contextual

—

SWE-bench Verified Score(%)

~70%

80.9%

Base AI Model

GPT-4, GPT-4o, o1, and newer models

—

Base Model Version

GPT-5.2

—

Context Window Size(tokens)

Up to 200,000 tokens (o1 models)

—

Context Window(tokens)

200K+ tokens

200K+

Custom GPT/Agent Creation

Yes; extensive custom GPT ecosystem

—

Third-Party App Integrations(count)

Thousands via Zapier and API

—

Ecosystem Integration

Standalone platform

—

Enterprise SLA & Compliance

Yes (ChatGPT Enterprise custom terms)

—

On-Premise Deployment

Limited (enterprise only)

—

Third-Party Integrations(score/10)

Extensive (1000+ integrations)

—

User Interface Rating(category)

Consumer-friendly, polished

—

Monthly Active Users(millions)

200+

Claude

Monthly Website Visits(millions)

1,500+

—

Launch Date

November 2022

—

Platform Availability

Web, iOS, Android, API

—

Free Tier Availability

Free basic features

—

Primary User Age Demographic(% under 35)

Mixed

—

Free Tier Message Limit(messages/month)

40 messages/month (ChatGPT Go)

—

Monthly Premium Cost(USD)

$20 (Plus) / $200 (Pro)

—

Cost for Annual Pro Usage(USD/year)

$240 (Plus annually)

—

Setup Time(days)

5-10 (account creation, app download)

—

Third-Party App Connections(apps)

Thousands via Zapier

—

Windows OS Integration

Separate application

—

Data Privacy Model

Processes public web data only

—

Developer Preference Rating(%)

Claude

—

Coding Task Excellence(benchmark)

Excellent (leading tool)

—

Reasoning Quality(benchmark)

Superior (Claude 4.6)

—

Company Valuation(billions USD)

Claude

—

ChatGPT vs Claude

ChatGPT

Claude

Short Answer

Our Verdict

Track this comparison

Key Differences at a Glance

Key Facts & Figures

Key Differences

Related Comparisons

Related Articles

Explore Entities

More Software

People Also Compare

Full Comparison

Visual Comparison

Pros & Cons

ChatGPT

Pros

Cons

Claude

Pros

Cons

Frequently Asked Questions

Resources & Learn More

Where to Buy

Wikipedia

Related Articles

Videos

SmartReview Ratings

Full Comparison

Visual Comparison

ChatGPT vs Claude

ChatGPT

Claude

Short Answer

Our Verdict

🔔Track this comparison

Key Differences at a Glance

Key Facts & Figures

Key Differences

Related Comparisons

Related Articles

Explore Entities

More Software

People Also Compare

Full Comparison

Visual Comparison

Pros & Cons

ChatGPT

Pros

Cons

Claude

Pros

Cons

Frequently Asked Questions

Resources & Learn More

Where to Buy

Wikipedia

Related Articles

Videos

SmartReview Ratings

Full Comparison

Visual Comparison

Track this comparison