deepseek Open weights Reasoning

R1 0528 pricing & benchmarks

Released May 28, 2025 · served by 4 providers

Input / 1M tokens$0.500
Output / 1M tokens$2.15
Context window164K
Cached input$0.350
1M in + 300K out$1.15

What R1 0528 is

May 28th update to the [original DeepSeek R1](/deepseek/deepseek-r1) Performance on par with [OpenAI o1](/openai/o1), but open-sourced and with fully open reasoning tokens. It's 671B parameters in size, with 37B active...

Providers and prices for R1 0528

The same model costs different amounts depending on who serves it. Prices are per 1M tokens, cheapest first. Uptime and throughput are OpenRouter's rolling measurements, not vendor claims.

ProviderInputOutputContext QuantUptime 24hThroughput
DeepInfracheapest $0.500 $2.15 164K fp4 99.9%
SiliconFlow $0.500 $2.18 164K fp8 97.9%
StreamLake $0.571 $2.29 128K 99.6%
Novita $0.700 $2.50 164K fp8 99.9%

R1 0528 benchmark results

Design Arena head-to-head results, as reported through the OpenRouter model API.

CategoryArenaRankEloWin rate
dataviz models #44 1208 60.7%
svg models #54 1088 48.7%
3d models #58 1169 53.4%
website models #64 1174 52.7%
codecategories models #65 1169 52.6%
gamedev models #65 1154 49.5%
uicomponent models #66 1141 55.1%

FAQ

How much does the R1 0528 API cost?

$0.500 per 1M input tokens and $2.15 per 1M output tokens. A typical workload of 1M input + 300K output tokens costs about $1.15.

Is R1 0528 free?

No. It is a paid model starting at $0.500 per 1M input tokens, though some providers offer trial credits.

Which provider is cheapest for R1 0528?

DeepInfra at $0.500 input / $2.15 output per 1M tokens (fp4 quantization).

Can I self-host R1 0528?

Yes — weights are published as deepseek-ai/DeepSeek-R1-0528 on Hugging Face, so you can run it on your own hardware.