What is the difference between Claude 3.7 Thinking Sonnet and GPT-4.1 Mini?

Claude 3.7 Thinking Sonnet is developed by Anthropic while GPT-4.1 Mini is developed by OpenAI. Claude 3.7 Thinking Sonnet has a 200K token context window vs GPT-4.1 Mini's 1.0M. in 1 community votes on Rival, Claude 3.7 Thinking Sonnet wins 100% of head-to-head matchups. These results are based on blind head-to-head voting across 54 challenges.

Which is better, Claude 3.7 Thinking Sonnet or GPT-4.1 Mini?

Based on 1 community votes on Rival, Claude 3.7 Thinking Sonnet wins 100% of head-to-head matchups against GPT-4.1 Mini.

How much does Claude 3.7 Thinking Sonnet cost compared to GPT-4.1 Mini?

Claude 3.7 Thinking Sonnet costs $6/M input tokens and GPT-4.1 Mini costs $0.4/M input tokens. GPT-4.1 Mini is $5.60/M cheaper per input. The more expensive model wins 100% of duels, so the premium may be justified by quality.

How are Claude 3.7 Thinking Sonnet vs GPT-4.1 Mini votes collected on Rival?

Rival presents both models' outputs side-by-side in blind duels. Voters see the responses but don't know which model produced each one, eliminating brand bias. 1 votes have been collected for this pair across 54 challenges. All vote data is part of Rival's open dataset.

Rival

Updated Apr 14, 2025

Claude 3.7 Thinking SonnetvsGPT-4.1 Mini

49 fights queued

Why Claude 3.7 Thinking Sonnet?

Dead even. This one's a coin flip.

Why GPT-4.1 Mini?

18.2x cheaper overall$0.40/M in · $1.60/M out

5.2x more context1.0M

2 months newerApr 2025

Claude 3.7 Thinking SonnetGPT-4.1 Mini

Input price

$6.00/M

$0.40/M

Output price

$30.00/M

$1.60/M

Context

200K

1.0M

Released

Feb 2025

Apr 2025

Favorites

Claude 3.7 Thinking SonnetGPT-4.1 Mini

Movie

The Shawshank Redemption

19948.7

Inception

20108.4

Album

Kind of Blue

Miles Davis

Dark Side Of The Moon

suisside

Book

Frankenstein or The Modern Prometheus

Mary Shelley

Sapiens

Yuval Noah Harari

City

Amsterdam

Netherlands

Tokyo

Japan

Game

Portal 2

Shooter, Puzzle4.6

The Legend of Zelda: Breath of the Wild

Adventure, Action4.5

12 / 49

Estimate Complexityreasoning

Tests an AI's ability to make educated estimates based on technical knowledge

Claude 3.7 Thinking Sonnet

Anthropic · Feb 2025text

Nothing here. The model returned empty. We stared at it for a while.

Try this prompt

GPT-4.1 Mini

Openai · Apr 2025text

Certainly! Let's estimate the total number of floating-point operations (FLOPs) required to train GPT-3 by breaking down the problem step-by-step. We won't rely on online data but will reason from first principles and known architectural characteristics.

Step 1: Understand GPT-3's basic specifications

GPT-3, the largest model version, has approximately:

Number of parameters (weights): 175 billion (1.75 × 10¹¹)
Number of layers: about 96 transformer layers
Model size: 175B parameters is widely known

Step 2: Estimate FLOPs per forward pass per token

A transformer forward pass consists mainly of:

Matrix multiplications in self-attention
Matrix multiplications in feedforward network (FFN)

Two main components dominate FLOPs:

Self-attention: roughly proportional to ( d_{model}^2 ) per token (where ( d_{model} ) is hidden size)
Feedforward layer: 2 linear layers per transformer layer, roughly (4 d_{model}^2) multiplications per token (due to expanding dimension by factor 4 in FFN)

Step 3: Extract hidden size and sequence length

From papers and common configurations:

Hidden size ( d_{model} approx 12,288 ) (for 175B model)
Sequence length used during training ( L approx 2048 ) tokens (max context length)

Step 4: Calculate approximate FLOPs per token per forward pass

Using formulas from transformer FLOPs estimation:

For each transformer layer, per token, forward pass FLOPs roughly:

[ ext{FLOPs}_ ext{layer/token} approx 2 imes d_{model}^2 ]

(for attention and FFN combined, includes multiplications and additions)

Because the FFN (feedforward network) is the heaviest part, it dominates, so let's assume:

( sim 8 imes d_{model}^2 ) FLOPs per token per layer (this is a common estimate for one forward pass)

Substituting:

[ 8 imes (12,288)^2 approx 8 imes 1.5 imes 10^8 = 1.2 imes 10^9 ext{ FLOPs per token per layer} ]

For 96 layers:

[ 1.2 imes 10^9 imes 96 = 1.15 imes 10^{11} ext{ FLOPs per token (forward pass)} ]

Step 5: Account for backward pass

Backpropagation generally costs about 2× to 3× the forward pass FLOPs because computing gradients involves forward pass and backward pass computations.

Use factor 3 to be safe:

[ ext{FLOPs} = 3 imes 1.15 imes 10^{11} = 3.45 imes 10^{11} ext{ FLOPs per token (forward + backward)} ]

Step 6: Calculate total number of tokens used in training

From publicly known info, GPT-3 was trained on about 300 billion tokens (around 3 × 10¹¹ tokens) of text.

Step 7: Calculate total FLOPs for full training

Multiply FLOPs per token by total tokens:

[ 3.45 imes 10^{11} imes 3 imes 10^{11} = 1.035 imes 10^{23} ext{ FLOPs} ]

Final estimation:

[ oxed{ ext{Total FLOPs to train GPT-3} approx 10^{23} quad ext{FLOPs} } ]

Summary:

Per token forward + backward FLOPs: (sim 3.5 imes 10^{11})
Number of training tokens: (sim 3 imes 10^{11})
Total training FLOPs: (sim 10^{23})

Additional notes:

This estimate matches the order of magnitude widely discussed in ML circles.
Variations depend on exact training steps, optimizer, precision (FP16/32), etc.
Might be a bit lower or higher depending on exact implementation.

So, the training of GPT-3 requires on the order of ( 10^{23} ) floating point operations.

Try this prompt

Vote

2 / 49

AI Board Game Logicreasoning

Tests an AI's ability to understand game rules and strategy

Claude 3.7 Thinking Sonnet

Anthropic · Feb 2025svg

Nothing here. The model returned empty. We stared at it for a while.

Try this prompt

GPT-4.1 Mini

Openai · Apr 2025svg

Try this prompt

Vote

11 / 49

Character Voice Testconversation

Tests an AI's ability to write in distinct character voices

Claude 3.7 Thinking Sonnet

Anthropic · Feb 2025text

Nothing here. The model returned empty. We stared at it for a while.

Try this prompt

GPT-4.1 Mini

Openai · Apr 2025text

Nothing here. The model returned empty. We stared at it for a while.

Try this prompt

Vote

This matchup has more rounds

37+ more head-to-head results. Free. Not a trick.

Free account. No card required. By continuing, you agree to Rival's Terms and Privacy Policy

Our Verdict

Claude 3.7 Thinking SonnetWinner

GPT-4.1 MiniRunner-up

Pick Claude 3.7 Thinking Sonnet. In 1 blind votes, Claude 3.7 Thinking Sonnet wins 100% of the time. That's not luck.

GPT-4.1 Mini is 19x cheaper per token — worth considering if cost matters.

Clear winner

Writing DNA

Style Comparison

Similarity

100%

Claude 3.7 Thinking Sonnet uses 2.6x more sentence length

Claude 3.7 Thinking Sonnet

GPT-4.1 Mini

61%Vocabulary59%

49wSentence Length19w

0.64Hedging0.46

2.2Bold4.7

4.5Lists4.9

0.00Emoji0.00

1.81Headings0.76

0.21Transitions0.17

Based on 21 + 23 text responses

Ask them anything yourself

279 AI models invented the same fake scientist.

We read every word. 250 models. 2.14 million words. This is what we found.

Free preview13 of 58 slides

Download the free preview or get all 58 slides for $49

FAQ

Common questions

Keep going

Claude 3.7 Thinking Sonnet vs Llama 4 MaverickNew provider

Claude 3.7 Thinking Sonnet vs Gemini 2.5 Pro Preview 06-05New provider

Claude 3.7 Thinking Sonnet vs MiniMax M2.5New provider

Step 1: Understand GPT-3's basic specifications

GPT-3, the largest model version, has approximately:

Number of parameters (weights): 175 billion (1.75 × 10¹¹)
Number of layers: about 96 transformer layers
Model size: 175B parameters is widely known

Step 2: Estimate FLOPs per forward pass per token

A transformer forward pass consists mainly of:

Matrix multiplications in self-attention
Matrix multiplications in feedforward network (FFN)

Two main components dominate FLOPs:

Self-attention: roughly proportional to ( d_{model}^2 ) per token (where ( d_{model} ) is hidden size)
Feedforward layer: 2 linear layers per transformer layer, roughly (4 d_{model}^2) multiplications per token (due to expanding dimension by factor 4 in FFN)

Step 3: Extract hidden size and sequence length

From papers and common configurations:

Hidden size ( d_{model} approx 12,288 ) (for 175B model)
Sequence length used during training ( L approx 2048 ) tokens (max context length)

Step 4: Calculate approximate FLOPs per token per forward pass

Using formulas from transformer FLOPs estimation:

For each transformer layer, per token, forward pass FLOPs roughly:

[ ext{FLOPs}_ ext{layer/token} approx 2 imes d_{model}^2 ]

(for attention and FFN combined, includes multiplications and additions)

Because the FFN (feedforward network) is the heaviest part, it dominates, so let's assume:

( sim 8 imes d_{model}^2 ) FLOPs per token per layer (this is a common estimate for one forward pass)

Substituting:

[ 8 imes (12,288)^2 approx 8 imes 1.5 imes 10^8 = 1.2 imes 10^9 ext{ FLOPs per token per layer} ]

For 96 layers:

[ 1.2 imes 10^9 imes 96 = 1.15 imes 10^{11} ext{ FLOPs per token (forward pass)} ]

Step 5: Account for backward pass

Backpropagation generally costs about 2× to 3× the forward pass FLOPs because computing gradients involves forward pass and backward pass computations.

Use factor 3 to be safe:

[ ext{FLOPs} = 3 imes 1.15 imes 10^{11} = 3.45 imes 10^{11} ext{ FLOPs per token (forward + backward)} ]

Step 6: Calculate total number of tokens used in training

From publicly known info, GPT-3 was trained on about 300 billion tokens (around 3 × 10¹¹ tokens) of text.

Step 7: Calculate total FLOPs for full training

Multiply FLOPs per token by total tokens:

[ 3.45 imes 10^{11} imes 3 imes 10^{11} = 1.035 imes 10^{23} ext{ FLOPs} ]

Final estimation:

[ oxed{ ext{Total FLOPs to train GPT-3} approx 10^{23} quad ext{FLOPs} } ]

Summary:

Per token forward + backward FLOPs: (sim 3.5 imes 10^{11})
Number of training tokens: (sim 3 imes 10^{11})
Total training FLOPs: (sim 10^{23})

Additional notes:

This estimate matches the order of magnitude widely discussed in ML circles.
Variations depend on exact training steps, optimizer, precision (FP16/32), etc.
Might be a bit lower or higher depending on exact implementation.

So, the training of GPT-3 requires on the order of ( 10^{23} ) floating point operations.

Claude 3.7 Thinking Sonnet vs GPT-4.1 Mini

Which is better, Claude 3.7 Thinking Sonnet or GPT-4.1 Mini?

Key Differences Between Claude 3.7 Thinking Sonnet and GPT-4.1 Mini

Why Claude 3.7 Thinking Sonnet?

Why GPT-4.1 Mini?

Favorites

Estimate Complexityreasoning

Claude 3.7 Thinking Sonnet

GPT-4.1 Mini

Step 1: Understand GPT-3's basic specifications

Step 2: Estimate FLOPs per forward pass per token

Step 3: Extract hidden size and sequence length

Step 4: Calculate approximate FLOPs per token per forward pass

Step 5: Account for backward pass

Step 6: Calculate total number of tokens used in training

Step 7: Calculate total FLOPs for full training

Final estimation:

Summary:

Additional notes:

AI Board Game Logicreasoning

Claude 3.7 Thinking Sonnet

GPT-4.1 Mini

Character Voice Testconversation

Claude 3.7 Thinking Sonnet

GPT-4.1 Mini

This matchup has more rounds

Style Comparison

Common questions

What is the difference between Claude 3.7 Thinking Sonnet and GPT-4.1 Mini?

Which is better, Claude 3.7 Thinking Sonnet or GPT-4.1 Mini?

How much does Claude 3.7 Thinking Sonnet cost compared to GPT-4.1 Mini?

How are Claude 3.7 Thinking Sonnet vs GPT-4.1 Mini votes collected on Rival?

Why Claude 3.7 Thinking Sonnet?

Why GPT-4.1 Mini?

Favorites

Estimate Complexityreasoning

Claude 3.7 Thinking Sonnet

GPT-4.1 Mini

Step 1: Understand GPT-3's basic specifications

Step 2: Estimate FLOPs per forward pass per token

Step 3: Extract hidden size and sequence length

Step 4: Calculate approximate FLOPs per token per forward pass

Step 5: Account for backward pass

Step 6: Calculate total number of tokens used in training

Step 7: Calculate total FLOPs for full training

Final estimation:

Summary:

Additional notes:

AI Board Game Logicreasoning

Claude 3.7 Thinking Sonnet

GPT-4.1 Mini

Character Voice Testconversation

Claude 3.7 Thinking Sonnet

GPT-4.1 Mini

This matchup has more rounds