DeepSeek-R1-0528: benchmarks, price and rank
DeepSeek-R1-0528 from DeepSeek is 114th of 123 rated AI models on Spezia's leaderboard. It costs $0.50 in, $2.15 out per million tokens.
- 114th of 123 rated models
- Rating 485
- Rated on 7 benchmarks
- Open weights
- Released May 28, 2025
- 164,000 tokens of context
Where it is strongest
Not in the top tenth on any benchmark in use.
Where it falls behind
- 56th of 56 on BrowseComp
- 66th of 69 on ARC-AGI-2
- 53rd of 56 on SWE-bench Verified
Price: $0.50 in, $2.15 out per million tokens.
Its headline results
The benchmarks that count most in the rating, with its place among the models measured the same way, and where each number came from.
- BrowseComp8.9%
Questions whose answers are hidden deep on the web and take many searches and pages to track down.
56th of 56. The maker's own number, from DeepSeek, via LLM Stats.
- Humanity's Last Exam17.7%
About 2,500 questions written by experts in more than a hundred fields to be as hard as they could make them, many needing a diagram or image.
56th of 74. The maker's own number, from DeepSeek, via LLM Stats.
- SimpleBench40.8%
Multiple-choice puzzles about everyday situations, space, time and people, written to trip up models that pattern-match instead of think. Ordinary adults do well on them.
63rd of 68. Independent result from Epoch AI.
- SWE-bench Verified44.6%
Real bug reports and feature requests from popular open-source Python projects. The model has to change the code so the project's hidden tests pass.
53rd of 56. The maker's own number, from DeepSeek, via LLM Stats.
- GPQA Diamond76.3%
Graduate-level questions in biology, physics and chemistry, written by PhDs so that searching the web does not give the answer away.
93rd of 103. Independent result from Epoch AI.
- ARC-AGI-21.1%
Visual puzzles of colored grids, each with a new rule to work out from a few examples.
66th of 69. Independent result from Epoch AI.
- MMLU-Pro85%
Twelve thousand multiple-choice questions across fourteen subjects, from law and medicine to engineering, with ten options each.
18th of 56. The maker's own number, from DeepSeek, via LLM Stats.
2 more results in its full profile.
Every result with its source, what it costs per task, what jumped with its release, side by side with any other model, and answers to your questions about when to use it.
Models ranked near it
How the rating works, and every benchmark it uses: the model leaderboard.