Claude 3.7 Thinking SonnetvsGemini 2.5 Pro Experimental
Claude 3.7 Thinking Sonnet wins 55% of 402 blind community votes against Gemini 2.5 Pro Experimental on Rival.
| Spec | Claude 3.7 Thinking Sonnet | Gemini 2.5 Pro Experimental |
|---|---|---|
| Win rate | 55% | 45% |
| Input price | $6/M tokens | $1/M tokens |
| Output price | $30/M tokens | $2/M tokens |
| Context window | 200K tokens | 1.0M tokens |
| Released | Feb 2025 | Mar 2025 |
| MMLU | 77.1% | 90.0% |
Our Verdict
Claude 3.7 Thinking SonnetWinner
Gemini 2.5 Pro ExperimentalRunner-up
Claude 3.7 Thinking Sonnet has the edge overall. In 402 blind votes, Claude 3.7 Thinking Sonnet wins 55% of the time.
Pick Claude 3.7 Thinking Sonnet for Web Design, Image Generation. Pick Gemini 2.5 Pro Experimental for Reasoning, Conversation. Gemini 2.5 Pro Experimental costs 15x less per token.
Slight edge
API pricing
Cost per 1M tokens
Claude 3.7 Thinking Sonnet
Input
$6.00
Output
$30.00
Gemini 2.5 Pro Experimental
Input
$1.00
6.0× cheaper
Output
$2.00
15× cheaper
Gemini 2.5 Pro Experimental is cheaper on both — 6.0× input, 15× output
Writing DNA
Style Comparison
Similarity
34%
Claude 3.7 Thinking Sonnet uses 4.6x more headings
Claude 3.7 Thinking Sonnet
Gemini 2.5 Pro Experimental
61%Vocabulary54%
49wSentence Length15w
0.64Hedging0.35
2.2Bold5.6
4.5Lists3.9
0.00Emoji0.00
1.81Headings0.39
0.21Transitions0.17
Based on 21 + 18 text responses
Research
What we learned reading every model
FAQ
Common questions
Keep exploring
More comparisons
Claude 3.7 Thinking Sonnet vs Ox AlphaNew provider
Claude 3.7 Thinking Sonnet vs MiniMax M3New provider
Claude 3.7 Thinking Sonnet vs GPT-5New provider
Claude 3.7 Thinking Sonnet vs Llama 4 MaverickNew provider
Claude 3.7 Thinking Sonnet vs GPT-5 CodexNew provider
Claude 3.7 Thinking Sonnet vs GPT-5 MiniNew provider
Claude 3.7 Thinking Sonnet vs GPT-5 NanoNew provider
Claude 3.7 Thinking Sonnet vs GPT-5 ProNew provider