AIbase
EN
Home
AI Models Finder
Deepseek R1 0528 (May '25)
D

Deepseek R1 0528 (May '25)

DeepSeek-R1-0528: The 685B open-source domestic inference flagship. Post-training strengthens the thought chain. It achieves 87.5% on AIME25, approaching o3. Hallucination is reduced by 50%. It supports tool calls and has dual contexts of 64K/128K. It is available for free commercial use under the MIT license.
Intelligence(Relatively Strong)
Speed(Slow)
Input Supported Modalities
No
Is Reasoning Model
128,000
Context Window
131,072
Maximum Output Tokens
-
Knowledge Cutoff
Go Compare

Pricing

￥3.6 /M tokens
Input
￥15.48 /M tokens
Output
￥6.91 /M tokens
Blended Price

Quick Simple Comparison

Input

Output

DeepSeek-Coder-V2
DeepSeek V3 0324 (Mar' 25)
DeepSeek V3 0324 (Mar '25)

Basic Parameters

DeepSeek R1 0528 (May '25)Technical Parameters
Parameter Count
671,000.0M
Context Length
128.00k tokens
Training Data Cutoff
Open Source Category
Open Weights (Permissive License)
Multimodal Support
Text Only
Throughput
45
Release Date
2025-05-28
Response Speed
23.333,052 tokens/s

Benchmark Scores

Below is the performance of DeepSeek R1 0528 (May '25) in various standard benchmark tests. These tests evaluate the model's capabilities in different tasks and domains.
Intelligence Index
68.29
Large Language Model Intelligence Level
Coding Index
58.66
Indicator of AI model performance on coding tasks
Math Index
93.8
Capability indicator in solving mathematical problems, mathematical reasoning, or performing math-related tasks
MMLU Pro
84.9
Massive Multitask Multimodal Understanding - Testing understanding of text, images, audio, and video
GPQA
81.3
Graduate Physics Questions Assessment - Testing advanced physics knowledge with diamond science-level questions
HLE
14.9
The model's comprehensive average score on the Hugging Face Open LLM Leaderboard
LiveCodeBench
77
Specific evaluation focused on assessing large language models' ability in real-world code writing and solving programming competition problems
SciCode
40.3
The model's capability in code generation for scientific computing or specific scientific domains
HumanEval
97
Score achieved by the AI model on the specific HumanEval benchmark test set
Math 500 Score
98.3
Score on the first 500 larger, more well-known mathematical benchmark tests
AIME Score
89.3
An indicator measuring an AI model's ability to solve high-difficulty mathematical competition problems (specifically AIME level)
AIbase
Empowering the Future, Your AI Solution Knowledge Base
English简体中文繁體中文にほんご
© 2025AIbase