Inference
EN
DeepSeek

DeepSeek-R1-0528

DeepSeek-R1-0528 is a reasoning model from DeepSeek in the DeepSeek-R1 family. 11 benchmarks count toward its index, in 4 categories.

  • Open weights
  • Reasoning
  • Input: text
  • Released 2025-05
RankedEnough results for an index.
IndexAll categories, out of 100.44.6±8.9Median 51.0
CoverageIndex weight that has results.60%
SpeedOutput tokens per second.20/sMedian 67/s
First tokenSeconds to the first output token.1.9 sMedian 1.3 s
Input / 1MUS dollars per 1M input tokens.$0.50Median $0.60
Output / 1MUS dollars per 1M output tokens.$2.15Median $2.55
ContextMaximum tokens in one request.164KMedian 262K
EloLMArena rating and rank.1428±6 #101Median 1422

Details

From OpenRouter
API IDThe model ID on OpenRouter.
deepseek/deepseek-r1-0528
Max outputMaximum output tokens per request.
33K
Knowledge cutoffLast date of training data.
31 Mar 2025
ToolsTool calls on OpenRouter.
Yes
JSON outputOutput follows a given JSON schema.
Yes
Cache read / 1MUS dollars per 1M cached tokens.
$0.35
WeightsLink to the published weights.
Hugging Face (opens in a new tab) ↗
RetiresDate OpenRouter removes the model.
—

Where to run

APIs that serve this model.
5 providers
ProviderTypeMaker, cloud, host or router.InputOutputContext
Deep InfraHost$0.50$2.15164K
OpenRouterRouter$0.50$2.15164K
SiliconFlowHost$0.50$2.18164K
IO.NETHost$0.524$2.21128K
Novita AIHost$0.70$2.50164K

US dollars per 1M tokens, list price. From models.dev (opens in a new tab) (MIT).

Run locally

684.53B parameters
16-bit
1438.5 GB
8-bit
764.7 GB
4-bit
436.7 GB
Weights
Hugging Face (opens in a new tab) ↗

Memory need with a short context. The 4-bit copy needs more than 128 GB. Compare local models

Capabilities

Index per category, out of 100.
Out of 100
AgenticMulti-step tasks with tools.None counted——
CodingCode writing and repair.2 counted46.9155 / 209
ReasoningLogic problems and puzzles.2 counted38.2173 / 227
MultimodalTasks with images and text.None counted——
KnowledgeFacts and expert knowledge.5 counted46.0165 / 268
MultilingualTasks in many languages.None counted——
InstructionTasks with strict prompt rules.None counted——
MathMath problems.2 counted45.9105 / 168

Results

11 counted
BenchmarkThe benchmark name.ResultThe score from the source.PlacePlace among results on this benchmark.
American Invitational Mathematics Examination 2024Math91.4%6 / 27
Massive Multitask Language Understanding ProfessionalKnowledge85.0%20 / 75
LiveCodeBench: Holistic and Contamination Free Evaluation of Large Language Models for CodeCoding73.3%4 / 12
American Invitational Mathematics Examination 2025Math87.5%22 / 53
MMLU-ReduxKnowledge93.4%8 / 19
Graduate-Level Google-Proof Q&AKnowledge81.0%64 / 141
Measuring Short-Form Factuality in Large Language ModelsKnowledge27.8%7 / 14
Software Engineering Benchmark VerifiedCoding57.6%89 / 114
ARC-AGI-1 (semi-private)Reasoning21.2%104 / 126
ARC-AGI-2 (semi-private)Reasoning1.1%111 / 131
Humanity's Last ExamKnowledge17.7%68 / 77
ARC-AGI-2 (public eval)Reasoning0.3%99 / 122
ARC-AGI-1 (public eval)Reasoning27.0%100 / 116

Sources

BenchLMCC BY-NC 4.0OpenRouterTerms restrict reuseARC Prize FoundationTerms restrict reusemodels.devMIT

History

5 releases
Index45.6 → 44.6
23 Oct 202612 Nov 2026
Output / 1M$2.15 → $2.15
23 Oct 202612 Nov 2026

Questions

How much does DeepSeek-R1-0528 cost?
$0.50 per 1M input tokens, $2.15 per 1M output tokens.
How good is DeepSeek-R1-0528?
Its index is 44.6 out of 100. That is place 180 of 283 ranked models.
What is DeepSeek-R1-0528 best at?
Knowledge: 46.0 out of 100, place 165 of 268.
How fast is DeepSeek-R1-0528?
20 output tokens per second, 1.9 s to the first token.
What is the context window of DeepSeek-R1-0528?
164K tokens.
Is DeepSeek-R1-0528 open source?
Yes. Its weights are published.
When was DeepSeek-R1-0528 released?
May 2025.

Badge

DeepSeek-R1-0528 on Inference360: place 180 of 283, index 44.6
HTML
<a href="https://inference360.ai/models/deepseek-r1-0528"><img src="https://inference360.ai/badges/deepseek-r1-0528.svg" alt="DeepSeek-R1-0528 on Inference360: place 180 of 283, index 44.6" height="20"></a>
Markdown
[![DeepSeek-R1-0528 on Inference360: place 180 of 283, index 44.6](https://inference360.ai/badges/deepseek-r1-0528.svg)](https://inference360.ai/models/deepseek-r1-0528)

Compare with

o4-mini44.7Gemma 4 26B A4B44.5DeepSeek V3.1 (Reasoning)43.0dots3-note Preview63.4

Lower price

dots3-note Preview63.4 · FreeApodex 1.1 Mini55.7 · FreeNemotron 3 Ultra50.8 · Free

More from DeepSeek

All models
DeepSeek V3 032437.1DeepSeek V3.1 (Reasoning)43.0DeepSeek V3.137.9DeepSeek V3.1 TerminusUnrankedR1 Distill Llama 70BUnranked