R1 0528
DeepSeek · released May 28, 2025 · deepseek-ai/DeepSeek-R1-0528
- Type
- Open weightsMIT License
- Params
- 685B
- Context
- 164K
37B active per word · about 123K words of context
Our take
Written Sep 30, 2026R1 0528 is a downloadable text model with a permissive licence, and its measured strength is competitive programming rather than general chat. It is a very large mixture-of-experts model, so running it yourself is a serious hardware undertaking.
Reach for it on competitive programming and algorithmic problem solving, where it scores 84.4% pass on LiveCodeBench, a contamination-free competitive programming benchmark, placing 4th of 16 on that board as of 28 Sep 2026. The licence allows commercial use, changes and redistribution (MIT), and you can try it through a host and download it later if it earns a place. Skip it if you need strong general chat, instruction following or creative writing, where its Arena placings are mid-field.
The case for it
- 84.4% pass on LiveCodeBench, placing 4th of 16 on that board as of 28 Sep 2026 — this measures set-piece algorithmic problems, not fixing issues in an existing codebase.
- The licence allows commercial use, changes and redistribution (MIT).
- You can download it and run it yourself, and hosted offers are listed across five hosts, so you can try it through a host and download it later if it earns a place.
The case against it
- Mid-field on general chat and instruction following: 79th of 168 on Arena Text (overall) and 91st of 168 on Arena Instruction Following as of 25 Sep 2026.
- 684.5 billion parameters in total, of which about 37 billion are used per token, so running it yourself needs substantial hardware.
How good is it?
EverydayGeneral questions and everyday reasoning
Arena Text (overall)79th of 168 · 1422
CodingWriting and fixing code on its own
Arena Coding80th of 168 · 1464
AgenticPlanning, calling tools, staying on task
Not yet scored on Arena Agent.
WritingDrafting and rewriting prose
Arena Creative Writing79th of 168 · 1389
Arena Creative Writing is the only board that has scored it for this.
These tests check whether a model follows instructions — a precondition for all the work above, but not a measure of how well that work is done.
Every published score for this model7 scoresEvery figure we hold, from 7 boards, with who ran it and a link to the source — including the boards no rating above is built on.
Can you run it yourself?
GeForce RTX 4090 · 24 GB
Too large for this card. The weights do not fit even with part of them offloaded to system memory.
Apple M1 Pro (16-core GPU) · 32 GB
Too large for this card. The weights do not fit even with part of them offloaded to system memory.
Radeon RX 7900 XT · 20 GB
Too large for this card. The weights do not fit even with part of them offloaded to system memory.
Memory use by level
Against a 24 GB card.
What is quantisation? →This model on every device we track71 devicesThe Q4 build most people download, on each device: what the weights come to, how much context the memory leaves, and whether it runs. Smallest device that runs it first.
Check against your own machine → · Where to rent it hosted →
Or rent it from someone else
Prices checked 4 hours ago — each listing carries its own date.
- per 1M tokens
- $0.50 in / $2.15 out
- Context served
- 164K
- Throughput
- Not measured
| Provider | In / out per 1M tokens | Context | Throughput | Trains on prompts | Logs prompts | Zero retention |
|---|---|---|---|---|---|---|
| DeepInfrafp4Direct and through OpenRouter | $0.50 / $2.15checked 4 hours ago | 164K33K max reply through OpenRouter | 22 tok/sthrough OpenRouter | DirectUnknownThrough OpenRouterNo | DirectUnknownThrough OpenRouterNo | DirectUnknownThrough OpenRouterConfirmed |
| OpenRouterOpenRouter's own listing | $0.50 / $2.15checked 4 hours ago | 164K | not measured | Unknown | Unknown | Unknown |
| SiliconFlowfp8Through OpenRouter | $0.50 / $2.18checked 4 hours ago | 164K147K max reply | 20 tok/s | No | No | Confirmed |
| StreamLakeThrough OpenRouter | $0.57 / $2.29checked 4 hours ago | 128K32K max reply | 53 tok/s | No | Yesunknown period | Unknown |
| Novita AIfp8Direct and through OpenRouter | $0.70 / $2.50checked 4 hours ago | 164K33K max reply through OpenRouter | 22 tok/sthrough OpenRouter | DirectUnknownThrough OpenRouterNo | DirectUnknownThrough OpenRouterNo | DirectUnknownThrough OpenRouterConfirmed |
DeepInfra
Novita AIAcross the 5 listings we hold: 4 say they do not train on prompts (2 of them only through OpenRouter), 0 say they do and 1 does not say. 3 appear in the zero-retention registry we check (2 of them only through OpenRouter); the rest are unknown to us.
What each host's API supports
From the parameter list each endpoint publishes. Streaming is omitted: nothing we hold reports it, for any model.
| Provider | Tool calling | JSON output | Strict schema |
|---|---|---|---|
| DeepInfrafp4Direct and through OpenRouter | ✗ | ✓ | ✓ |
| OpenRouterOpenRouter's own listing | ✓ | ✓ | ✓ |
| SiliconFlowfp8Through OpenRouter | ✓ | ✓ | ✓ |
| StreamLakeThrough OpenRouter | ✗ | ✓ | ✓ |
| Novita AIfp8Direct and through OpenRouter | ✓ | ✓ | ✗ |
Tool calling: 3 of 5 listings say yes, 2 say no. JSON output: 5 of 5 listings say yes. Strict schema: 4 of 5 listings say yes, 1 says no.
When we formed this view
Recent changes
What moved
first indexed by our pipelineEach date is the day we first saw the change, or the day the maker announced it.
What we do not know about this model yet
- We hold no measured file for it, so all 3 sizes on this page are calculated from the parameter count.
- Nothing we hold says whether an endpoint streams, so we do not show it either way.
- 1 of 5 listings does not say whether it trains on prompts, and 2 answer only through OpenRouter, not for their own listing.
- We hold no batch or off-peak rate for any of its listings.
- We hold a decode speed for it, but no prompt-processing (prefill) figure, so how long the input side of a job takes is unknown to us.
Licence and identifiers
What the licence allowsMIT License, what it allows commercially, and the identifiers you need to pull this model — its Hugging Face repo, our slug and a machine-readable card.
Licence
MIT License
Fully permissive: do anything with attribution. No patent grant, unlike Apache-2.0.
Identifiers
- Hugging Face
- deepseek-ai/DeepSeek-R1-0528
- Architecture
- Mixture of experts
- Takes in, gives back
- Text in, text out
- Catalogue slug
- deepseek-r1-0528