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 Aug 2, 2026R1 0528 is a large downloadable reasoning model from DeepSeek with a permissive MIT licence and a 163,840-token request limit. It excels at coding and hard-prompt tasks, though its creative writing and instruction-following scores lag well behind its own coding peak.
Choose this for coding workloads where measured ability matters, or for hard-prompt tasks that need a wide context window. It suits teams who want open-weight deployment with a permissive licence, and those who can shop around for the cheapest host. Skip it if you need strong creative writing, if consistent throughput across providers matters, or if you want a model whose maths score matches its coding peak.
The case for it
- Arena Coding Elo 1464.4 — its highest category, 42.3 points above its own general text rating.
- Permissive MIT licence allows commercial use, modification and redistribution.
- 163,840-token request limit, wide enough for long-document tasks.
- Significant price spread between hosts creates genuine budget options.
The case against it
- Creative writing and instruction following lag its coding peak by 70.8 and 67.1 Elo points respectively.
- Throughput varies dramatically by provider: 74 tokens per second at one host versus 18 at another, with four hosts not disclosing speed at all.
- Arena Maths Elo 1411.6 sits 10.5 points below its own overall text rating and 52.8 below its coding score.
How good is it?
IntelligencePuzzles, maths, exam questions
Arena Text (overall)60th of 143 · 1422.1
CodingWriting and fixing code on its own
Arena Coding62nd of 143 · 1464.5
AgenticPlanning, calling tools, staying on task
Nobody we watch has scored R1 0528 for this. We would take the rating from Arena Agent (IPS).
WritingWe do not rate this
Two boards come close and neither tests writing: Arena Creative Writing asks people which of two replies they prefer, and LiveBench Language tests whether a model understood a passage. So we show where R1 0528 placed and give it no mark out of five.
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, which is why they get no rating.
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?
- Fits in memory
- weights load entirely on the card
- Spills to system RAM
- some weights offload; much slower
- Too large
- will not load even with offload
- est
- size is calculated; the verdict could change by 10%
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.
All 3 sizes here are calculated, not measured. We hold no measured file for this model, so each size comes from the parameter count and every verdict above inherits that uncertainty.
Check against your own machine → · All 71 devices, with every size →
Or rent it from someone else
Cheapest of 8 live listings. Picked at the widest standard context we hold, within one quantisation slice, so the numbers beside it are a price one host actually charges.
- 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 |
|---|---|---|---|---|---|---|
| DeepInfrafp4 | $0.50 / $2.15 | 164K | not measured | Unknown | Unknown | Unknown |
| DeepInfrafp4 | $0.50 / $2.15 | 164K | 20 tok/s | No | No | Confirmed |
| OpenRouter | $0.50 / $2.15 | 164K | not measured | Unknown | Unknown | Unknown |
| SiliconFlowfp8 | $0.50 / $2.18 | 164K | 17 tok/s | No | No | Confirmed |
| StreamLake | $0.57 / $2.29 | 128K | 35 tok/s | No | Yesunknown period | Unknown |
| Novita AIfp8 | $0.70 / $2.50 | 164K | 20 tok/s | No | No | Confirmed |
| Novita AI | $0.70 / $2.50 | 164K | not measured | Unknown | Unknown | Unknown |
| Microsoft Azure AI | $1.49 / $5.94 | 164K | 69 tok/s | No | No | Confirmed |
DeepInfra
Novita AI
Microsoft Azure AIAcross the 8 listings we hold: 5 say they do not train on prompts, 0 say they do and 3 do not say. 4 appear in the zero-retention registry we check; the rest are unknown to us rather than confirmed either way.
What each host's API supports
From the parameter list each endpoint publishes. Streaming is omitted: nothing we hold reports it, for any model.
- ✓
- Supported
- ✗
- Not supported
- Not published
- host gave no parameter list
| Provider | Tool calling | JSON output | Strict schema |
|---|---|---|---|
| DeepInfrafp4 | |||
| DeepInfrafp4 | ✗ | ✓ | ✓ |
| OpenRouter | ✓ | ✓ | ✓ |
| SiliconFlowfp8 | ✓ | ✓ | ✓ |
| StreamLake | ✗ | ✓ | ✓ |
| Novita AIfp8 | ✓ | ✓ | ✓ |
| Novita AI | |||
| Microsoft Azure AI | ✗ | ✗ | ✗ |
Tool calling: 3 of 8 listings say yes, 3 say no, 2 publish no parameter list. JSON output: 5 of 8 listings say yes, 1 says no, 2 publish no parameter list. Strict schema: 5 of 8 listings say yes, 1 says no, 2 publish no parameter list.
Models people weigh against R1 0528
When we formed this view
Dates behind this page
Prices last checked 14h ago
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.
- 2 of 8 listings publish no parameter list, so what their API accepts is unknown to us.
- Nothing we hold says whether an endpoint streams, so we do not show it either way.
- 3 of 8 listings do not say whether they train on prompts.
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
- Modality record
- text->text
- Catalogue slug
- deepseek-r1-0528