Model Comparison

Comprehensive side-by-side analysis of model capabilities and performance

DeepSeek-R1-0528

DeepSeek

DeepSeek-R1-0528 is a language model developed by DeepSeek. It achieves strong performance with an average score of 62.2% across 22 benchmarks. It excels particularly in MMLU-Redux (93.4%), SimpleQA (92.3%), AIME 2024 (91.4%). The model shows particular specialization in factuality tasks with an average performance of 92.9%. It supports a 262K token context window for handling large documents. The model is available through 3 API providers. It's licensed for commercial use, making it suitable for enterprise applications. Released in 2025, it represents DeepSeek's latest advancement in AI technology.

DeepSeek-V3.2-Exp

DeepSeek

DeepSeek-V3.2-Exp is a language model developed by DeepSeek. It achieves strong performance with an average score of 66.1% across 14 benchmarks. It excels particularly in SimpleQA (97.1%), AIME 2025 (89.3%), MMLU-Pro (85.0%). It supports a 229K token context window for handling large documents. The model is available through 2 API providers. It's licensed for commercial use, making it suitable for enterprise applications. Released in 2025, it represents DeepSeek's latest advancement in AI technology.

DeepSeek-R1-0528

DeepSeek

2025-05-28

DeepSeek-V3.2-Exp

DeepSeek

2025-09-29

4 months newer

Pricing Comparison

Cost per million tokens (USD)

DeepSeek-R1-0528

Input:$0.50

Output:$2.15

DeepSeek-V3.2-Exp

$1.97 cheaper

Input:$0.27

Output:$0.41

Performance Metrics

Context window and performance specifications

DeepSeek-R1-0528

Larger context

Max Context:262.1K

Parameters:671.0B

DeepSeek-V3.2-Exp

Max Context:229.4K

Parameters:685.0B

Average performance across 21 common benchmarks

DeepSeek-R1-0528

+22.4%

Average Score:63.7%

DeepSeek-V3.2-Exp

Average Score:41.3%

Performance comparison across key benchmark categories

DeepSeek-R1-0528

factuality

92.9%

math

84.6%

code

53.7%

general

58.8%

agents

40.5%

DeepSeek-V3.2-Exp

factuality

+4.2%

97.1%

math

+1.8%

86.5%

code

+10.5%

64.3%

general

+3.3%

62.1%

agents

+3.5%

44.0%

Provider Availability & Performance

Available providers and their performance metrics

DeepSeek-R1-0528

3 providers

DeepInfra

Throughput: 45.04 tok/s

Latency: 0.61ms

Novita

Throughput: 37.96 tok/s

Latency: 1.18ms

DeepSeek

Throughput: 9 tok/s

Latency: 0.3ms

DeepSeek-V3.2-Exp

2 providers

Novita

ZeroEval

Throughput: 100 tok/s

Latency: 0.7ms

DeepSeek-R1-0528

+22.4%

Avg Score:63.7%

Providers:3

DeepSeek-V3.2-Exp

Avg Score:41.3%

Providers:2