Sulat.com
AI models
$10 off the fastest DeepSeek V4.1 Flash, Kimi K3 and GLM 5.3
$10 off the fastest DeepSeek V4.1 Flash, Kimi K3 and GLM 5.3 from Synthetic
Jiekou.AI logo

Model details

DeepSeek R1 0528

DeepSeek R1 0528 is a May 2025 update to the original DeepSeek R1, positioned as the strongest open-weight reasoning model in its family and described as approaching the performance tier of leading closed frontier models on reasoning benchmarks. The architecture follows a Mixture-of-Experts design with 671 billion total parameters and roughly 37 billion active per inference pass, balancing large-scale capacity with efficient routing for complex reasoning workloads. Its open-source weights are published on Hugging Face, and it exposes fully open reasoning tokens, giving downstream developers visibility into the chain-of-thought process that drives its answers.

The release focused on practical gains rather than a redesign: improved benchmark scores, enhanced front-end code generation, reduced hallucinations, and native support for JSON output and function calling. These refinements make it well suited for agent-style workflows that demand reliable tool use and structured responses, as well as for mathematics, logic, and programming tasks where the original R1 already excelled. With a 164K context window and long-form reasoning capability, R1 0528 fits production deployments where teams want frontier-style reasoning quality without relinquishing control over model weights or inference transparency.

Jiekou.AIdeepseek/deepseek-r1-0528deepseek-thinking
Open provider pageProvider docs

Quick Info

Powered by
Provider
Jiekou.AI
Model key
deepseek/deepseek-r1-0528
Release date
Jan 1, 2026
Last updated
Jan 1, 2026
Input modalities
Output modalities
Capabilities

Cost

Input token cost
$0.70
Output token cost
$2.50

Limits

Output tokens
32,768 tokens
Context window
163,840 tokens

Transparent token rates

Compare DeepSeek R1 0528 pricing

Rates are shown per one million tokens. Combined means one million input plus one million output tokens.

Browse this family
News search byUse it for free

Latest news about DeepSeek R1 0528

Cortecs

CoverageAnalysis

A third-party deep dive analyzes DeepSeek-R1-0528 as more than a cosmetic patch, detailing that it retains the Mixture-of-Experts architecture scaled up with 128K context (extendable via RoPE scaling) while delivering large mathematical-reasoning improvements. The post highlights the AIME 2025 jump from 70% to 87.5% as The piece contextualizes R1-0528 within DeepSeek's open-source trajectory and notes improvements relevant to vibe coding and function calling. It frames the release as a stealth incremental update that nonetheless shifts the competitive balance against closed frontier models in mathematical and multi-step reasoning wor

Jiekou.AI

Coverage

Last updated: May 12, 2026

Jiekou.AI

Coverage

The Chinese start-up DeepSeek has updated its R1 model, improving its performance in reasoning, logic, mathematics, and programming. This update, which reduces...

Jiekou.AI

CoverageBenchmark

DeepSeek’s New R1–0528: Performance Analysis and Benchmark Comparisons TL;DR: The newest DeepSeek R1 model is the most powerful among open-weight models, approaching performance of the leading …

Jiekou.AI

Coverage

The official DeepSeek API Change Log documents that as of 2026-04-24, the DeepSeek API supports DeepSeek-V4-Pro and DeepSeek-V4-Flash via both the OpenAI ChatCompletions and Anthropic interfaces, with the base URL unchanged. The model parameter should be set to `deepseek-v4-pro` or `deepseek-v4-flash` to access these n Critically, the legacy model names `deepseek-chat` and `deepseek-reasoner` are scheduled for discontinuation on 2026-07-24 — only ten days from the current date. During the transitional period, these aliases point to the non-thinking mode and thinking mode of `deepseek-v4-flash`, respectively. Because Jiekou.AI's DeepS

Videos about DeepSeek R1 0528

More models around DeepSeek R1 0528