LLM ReferenceLLM Reference

DeepSeek R1 0528

deepseek-r1-0528

Researched 29d ago

Last refreshed 2026-05-16. Next refresh: weekly.

Open SourceCodingRAGAgentsLong contextClassificationJSON / Tool use

DeepSeek R1 0528 is worth evaluating for coding, rag, and agents when its provider route and context window match the workload.

Decision context: Coding task fit, 6 tracked provider routes, and research from 2026-04-19.

Use it for

  • Teams evaluating coding, rag, and agents
  • Workloads that can use a 160K context window
  • Buyers comparing 4 tracked provider routes

Do not use it for

  • Vision or document-understanding workloads

Cheapest output

$0.300

Novita AI per 1M tokens

Provider routes

6

Tracked API hosts

Quality / dollar

Grade B

Ranked by benchmark score divided by cheapest output price

Freshness

2026-04-19

Researched 29d ago

fresh

Top use-case fit

Coding

Q/$ B

1 relevant benchmark in the decision map.

RAG

Included by capability and metadata signals in the decision map.

Agents

Included by capability and metadata signals in the decision map.

Provider price ladder

Compare all 6
ProviderInput / 1MOutput / 1MRoute
Novita AI$0.100$0.300
Serverless
Fireworks AI$0.560$1.68
Serverless
OpenRouter$0.500$2.15
Serverless
GCP Vertex AI$1.35$5.40
Serverless

Benchmark peer barsfor Coding

Migration checks

No linked migration route is available for this model yet.

About

DeepSeek R1 0528 has a 160K-token context window.

DeepSeek R1 0528 input tokens at $0.1/1M, output at $0.3/1M.

Capabilities

ReasoningStructured OutputsCode Execution

Benchmark Scores(2)

Scores are benchmark-specific and are direction-aware: the same numeric gap can mean very different outcomes across suites. Use the leaderboard context and this model's provider route to decide whether the winning margin is meaningful for your workload.
BenchmarkScoreVersionSource
Aider Polyglot71.42026-04https://aider.chat/docs/leaderboards
Google-Proof Q&A81.0diamondhttps://huggingface.co/deepseek-ai/DeepSeek-R1-0528

Rankings

Specifications

Released2025-01-01
Parameters671B
Context160K
ArchitectureDecoder Only
Specializationgeneral
Trainingpretrained

Created by

Advancing artificial general intelligence (AGI).

Hangzhou, Zhejiang, China
Founded 2023
Website