DeepSeek R1 benchmarks and performance, in charts
Every chart we tracked that shows DeepSeek R1, from the lab itself and from the trackers people trust, with how far each one travelled.
DeepSeek R1 on its own benchmark table
Drawn from the numbers huggingface.co published as text (DeepSeek-R1-Evaluation, read Mar 27, 2025). Each panel is one test; the highlighted bar is DeepSeek R1. Source: huggingface.co.
See the numbers
| English | OpenAI o1-1217 91.8, DeepSeek R1 90.8, DeepSeek V3 88.5, Claude-3.5-Sonnet-1022 88.3, GPT-4o 0513 87.2, OpenAI o1-mini 85.2 |
|---|---|
| Code | DeepSeek R1 65.9, OpenAI o1-1217 63.4, OpenAI o1-mini 53.8, GPT-4o 0513 34.2, Claude-3.5-Sonnet-1022 33.8 |
| Math | DeepSeek R1 79.8, OpenAI o1-1217 79.2, OpenAI o1-mini 63.6, DeepSeek V3 39.2, Claude-3.5-Sonnet-1022 16, GPT-4o 0513 9.3 |
| Chinese | DeepSeek R1 92.8, DeepSeek V3 90.9, OpenAI o1-mini 89.9, GPT-4o 0513 87.9, Claude-3.5-Sonnet-1022 85.4 |
Official charts from DeepSeek
As published on X, Hugging Face, GitHub and their blog. Each card opens our own page for the chart.
From benchmark trackers and others
More charts that feature DeepSeek R1
Benchmark trackers, other labs and researchers. Each card opens our own page for the chart.
Free with your email
See all 38 DeepSeek R1 charts
Every official launch chart and independent benchmark we found for DeepSeek R1, with who posted each one. Plus Monday's email: the week's charts that beat their usual likes.
Unlocks all 35 giants. One email a week, leave any time. What you get · Log in
More charts that feature DeepSeek R1 more
Make it yours
Your numbers, in a look-alike of the original.
Starts from the real numbers of "Model Performance Comparison Across Benchmarks" by github:zai-org/GLM-4. Replace every number with yours before you post. Make one like this, step by step
Look-alike score 9.0 of 10. Free pictures carry a small "Made with ViralChart" line. That line is how other people find us.
Questions people ask
How does DeepSeek R1 perform?
We hold 12 charts that show DeepSeek R1, posted between Jan 2025 and Mar 2026, 0 of them from DeepSeek. The most seen is "AI model benchmark scores" by Artificial Analysis with 285,814 views.
DeepSeek R1 vs DeepSeek V3: which is better?
1 charts put them side by side. Which wins depends on the benchmark; each chart shows its own test and who published it.
Who publishes DeepSeek R1 benchmark charts?
Artificial Analysis, SemiAnalysis, Arena, Statista, AMD, DeepSeek, @SemiAnalysis_, @arena, @Alibaba_Qwen, @AnthropicAI