DeepSeek R1 vs o3
Side-by-side benchmark comparison across coding, math, reasoning, speed, and pricing.
o3 by OpenAI wins on 18 of 22 benchmarks against DeepSeek R1 by DeepSeek, which leads on 3. 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, o3 scores 1380 on Chatbot Arena ELO compared to DeepSeek R1's 1355, while both score 84.0% on MMLU-Pro, while o3 scores 9.5 on MT-Bench compared to DeepSeek R1's 8.7, while o3 scores 92.0% on IFEval compared to DeepSeek R1's 87.8%, while o3 scores 49.4% on SimpleQA compared to DeepSeek R1's 30.1%, while o3 scores 86.0% on TruthfulQA compared to DeepSeek R1's 78.0%.
Coding: In coding, o3 scores 90.0% on HumanEval+ compared to DeepSeek R1's 86.0%, while o3 scores 69.1% on SWE-bench Verified compared to DeepSeek R1's 49.2%, while o3 scores 70.0% on LiveCodeBench compared to DeepSeek R1's 50.0%.
Math: In math, DeepSeek R1 scores 97.3% on MATH compared to o3's 96.0%, while o3 scores 98.5% on GSM8K compared to DeepSeek R1's 97.0%, while o3 scores 98.4% on AIME 2025 compared to DeepSeek R1's 79.8%.
Reasoning: In reasoning, o3 scores 83.3% on GPQA Diamond compared to DeepSeek R1's 71.5%, while o3 scores 75.7% on ARC-AGI compared to DeepSeek R1's 14.0%, while o3 scores 95.8% on BBH compared to DeepSeek R1's 89.0%, while o3 scores 94.0% on Winogrande compared to DeepSeek R1's 88.0%.
Context: In context, o3 scores 200K on Context Length compared to DeepSeek R1's 64K.
Multimodal: In multimodal, o3 scores 86.8% on MathVista compared to DeepSeek R1's 73.6%.
Pricing Comparison
DeepSeek R1 costs $0.70/1M input tokens and $2.5/1M output tokens, while o3 costs $2.0/1M input and $8.0/1M output. DeepSeek R1 is the more affordable option for API usage.
Speed Comparison
DeepSeek R1 generates output at 35 tok/s compared to o3's 40 tok/s, and the time to first token is 950 ms for DeepSeek R1 versus 800 ms for o3. o3 delivers faster throughput.
Verdict
For developers prioritizing math and affordability, DeepSeek R1 has the edge. For those who value coding and general intelligence and speed, o3 is the stronger choice.
View Individual Model Pages
DeepSeek R1 vs o3 — FAQ
Which is better, DeepSeek R1 or o3?
o3 wins on more benchmarks overall (18 vs 3). However, the best choice depends on your specific needs — each model excels in different areas.
How does DeepSeek R1 compare to o3 for coding?
o3 is better for coding, scoring 69.1% on SWE-bench Verified compared to 49.2%. SWE-bench tests real-world software engineering by resolving actual GitHub issues.
Is DeepSeek R1 cheaper than o3?
Yes, DeepSeek R1 is cheaper. DeepSeek R1 costs $0.70/1M input and $2.5/1M output tokens. o3 costs $2.0/1M input and $8.0/1M output tokens.
Which is faster, DeepSeek R1 or o3?
o3 is faster, generating output at 40 tok/s compared to 35 tok/s. Faster output speed means shorter wait times for API responses.
What benchmarks does the DeepSeek R1 vs o3 comparison cover?
This comparison covers 22 benchmarks including Chatbot Arena ELO, MMLU-Pro, HumanEval+, MT-Bench, IFEval, SimpleQA, MATH, TruthfulQA, and more. Metrics span general intelligence, coding, math, reasoning, speed, and cost categories.