Claude 3 Opus vs Claude Opus 4
Side-by-side benchmark comparison across coding, math, reasoning, speed, and pricing.
Claude Opus 4 by Anthropic wins on 12 of 15 benchmarks against Claude 3 Opus by Anthropic, which leads on 0. This head-to-head comparison covers coding, math, reasoning, speed, and pricing metrics from our benchmark data.
Category-by-Category Breakdown
General Intelligence: In general intelligence, Claude Opus 4 scores 1375 on Chatbot Arena ELO compared to Claude 3 Opus's 1240, while Claude Opus 4 scores 83.0% on MMLU-Pro compared to Claude 3 Opus's 68.0%, while Claude Opus 4 scores 88.0% on IFEval compared to Claude 3 Opus's 82.0%.
Coding: In coding, Claude Opus 4 scores 90.0% on HumanEval+ compared to Claude 3 Opus's 78.0%, while Claude Opus 4 scores 72.5% on SWE-bench Verified compared to Claude 3 Opus's 22.0%, while Claude Opus 4 scores 60.0% on LiveCodeBench compared to Claude 3 Opus's 32.0%.
Math: In math, Claude Opus 4 scores 86.0% on MATH compared to Claude 3 Opus's 68.0%, while Claude Opus 4 scores 95.0% on GSM8K compared to Claude 3 Opus's 88.0%.
Reasoning: In reasoning, Claude Opus 4 scores 79.6% on GPQA Diamond compared to Claude 3 Opus's 59.4%, while Claude Opus 4 scores 45.0% on ARC-AGI compared to Claude 3 Opus's 8.0%.
Context: In context, both score 200K on Context Length.
Pricing Comparison
Claude 3 Opus costs $15.0/1M input tokens and $75.0/1M output tokens, while Claude Opus 4 costs $15.0/1M input and $75.0/1M output. Both models are priced identically for input tokens.
Speed Comparison
Claude 3 Opus generates output at 25 tok/s compared to Claude Opus 4's 50 tok/s, and the time to first token is 800 ms for Claude 3 Opus versus 550 ms for Claude Opus 4. Claude Opus 4 delivers faster throughput.
Verdict
Claude Opus 4 leads across the board in coding, general intelligence, math, speed, making it the stronger overall choice in this comparison.
View Individual Model Pages
Claude 3 Opus vs Claude Opus 4 — FAQ
Which is better, Claude 3 Opus or Claude Opus 4?
Claude Opus 4 wins on more benchmarks overall (12 vs 0). However, the best choice depends on your specific needs — each model excels in different areas.
How does Claude 3 Opus compare to Claude Opus 4 for coding?
Claude Opus 4 is better for coding, scoring 72.5% on SWE-bench Verified compared to 22.0%. SWE-bench tests real-world software engineering by resolving actual GitHub issues.
Is Claude 3 Opus cheaper than Claude Opus 4?
Both models are priced the same at $15.0/1M input tokens. Claude 3 Opus output costs $75.0/1M and Claude Opus 4 output costs $75.0/1M.
Which is faster, Claude 3 Opus or Claude Opus 4?
Claude Opus 4 is faster, generating output at 50 tok/s compared to 25 tok/s. Faster output speed means shorter wait times for API responses.
What benchmarks does the Claude 3 Opus vs Claude Opus 4 comparison cover?
This comparison covers 15 benchmarks including Chatbot Arena ELO, MMLU-Pro, HumanEval+, IFEval, MATH, GPQA Diamond, SWE-bench Verified, Output Speed, and more. Metrics span general intelligence, coding, math, reasoning, speed, and cost categories.