Skip to content
Get an API key

Live · serverless · 2 regions

DeepSeek R1 0528

by DeepSeek

A reasoning model that writes out its chain of thought before answering, suited to math, logic and hard debugging.

  • reasoning
  • INT4
Serverless priceUSD
Input
$0.70per 1M tokens
Output
$2.40per 1M tokens

Billed per token, no minimums.

Reference

Specifications

Model ID
deepseek-r1-0528
Endpoint
POST /v1/chat/completions
Publisher
DeepSeek
Tasks
reasoning
Served context
128K tokens
Quantization
INT4
Decode speed
48 tok/smedian, one request
p50 first token
520 mslast 24 hours
Regions
us-east · us-west
Dedicated
On request

Last 24 hours

Latency and speed

Median of each 15-minute step on the shared pool, measured at the API edge. Time to first token includes queueing; decode speed is for one request.

Time to first token · ms

520 msmedian

p95 693 ms

Decode speed · tok/s

48 tok/smedian

p5 40 tok/s

Show the data
Median of each 15-minute step over the last 24 hours, newest first
StepTime to first tokenDecode speed
15m ago552 ms48 tok/s
30m ago530 ms47 tok/s
45m ago537 ms48 tok/s
1h ago517 ms49 tok/s
1h 15m ago530 ms46 tok/s
1h 30m ago527 ms46 tok/s
1h 45m ago518 ms49 tok/s
2h ago560 ms45 tok/s
2h 15m ago521 ms47 tok/s
2h 30m ago550 ms47 tok/s
2h 45m ago553 ms46 tok/s
3h ago562 ms46 tok/s
3h 15m ago537 ms47 tok/s
3h 30m ago551 ms48 tok/s
3h 45m ago525 ms48 tok/s
4h ago546 ms48 tok/s
4h 15m ago555 ms47 tok/s
4h 30m ago563 ms46 tok/s
4h 45m ago553 ms47 tok/s
5h ago513 ms48 tok/s
5h 15m ago547 ms48 tok/s
5h 30m ago523 ms48 tok/s
5h 45m ago520 ms48 tok/s
6h ago522 ms47 tok/s
6h 15m ago531 ms46 tok/s
6h 30m ago544 ms47 tok/s
6h 45m ago508 ms50 tok/s
7h ago731 ms37 tok/s
7h 15m ago527 ms48 tok/s
7h 30m ago510 ms47 tok/s
7h 45m ago550 ms48 tok/s
8h ago550 ms48 tok/s
8h 15m ago543 ms48 tok/s
8h 30m ago537 ms46 tok/s
8h 45m ago534 ms47 tok/s
9h ago504 ms49 tok/s
9h 15m ago520 ms49 tok/s
9h 30m ago519 ms49 tok/s
9h 45m ago516 ms48 tok/s
10h ago491 ms49 tok/s
10h 15m ago493 ms48 tok/s
10h 30m ago521 ms47 tok/s
10h 45m ago712 ms40 tok/s
11h ago486 ms49 tok/s
11h 15m ago520 ms47 tok/s
11h 30m ago527 ms49 tok/s
11h 45m ago487 ms49 tok/s
12h ago492 ms50 tok/s
12h 15m ago492 ms49 tok/s
12h 30m ago490 ms50 tok/s
12h 45m ago507 ms49 tok/s
13h ago481 ms50 tok/s
13h 15m ago504 ms49 tok/s
13h 30m ago516 ms48 tok/s
13h 45m ago477 ms50 tok/s
14h ago509 ms49 tok/s
14h 15m ago714 ms39 tok/s
14h 30m ago487 ms48 tok/s
14h 45m ago500 ms48 tok/s
15h ago482 ms51 tok/s
15h 15m ago502 ms49 tok/s
15h 30m ago491 ms49 tok/s
15h 45m ago488 ms51 tok/s
16h ago502 ms50 tok/s
16h 15m ago502 ms50 tok/s
16h 30m ago477 ms50 tok/s
16h 45m ago490 ms50 tok/s
17h ago519 ms48 tok/s
17h 15m ago478 ms49 tok/s
17h 30m ago497 ms49 tok/s
17h 45m ago503 ms48 tok/s
18h ago511 ms48 tok/s
18h 15m ago518 ms48 tok/s
18h 30m ago512 ms48 tok/s
18h 45m ago508 ms48 tok/s
19h ago485 ms50 tok/s
19h 15m ago500 ms49 tok/s
19h 30m ago537 ms48 tok/s
19h 45m ago509 ms47 tok/s
20h ago525 ms47 tok/s
20h 15m ago520 ms49 tok/s
20h 30m ago538 ms48 tok/s
20h 45m ago542 ms47 tok/s
21h ago497 ms50 tok/s
21h 15m ago693 ms40 tok/s
21h 30m ago521 ms48 tok/s
21h 45m ago501 ms50 tok/s
22h ago703 ms39 tok/s
22h 15m ago550 ms48 tok/s
22h 30m ago544 ms46 tok/s
22h 45m ago553 ms46 tok/s
23h ago532 ms48 tok/s
23h 15m ago533 ms49 tok/s
23h 30m ago553 ms46 tok/s
23h 45m ago544 ms48 tok/s
24h ago536 ms46 tok/s

API

Make a request

The same request shape as every other model on /v1/chat/completions. Switching between them is a change to one string.

curl https://api.weir.example/v1/chat/completions \
  -H "Authorization: Bearer $WEIR_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "deepseek-r1-0528",
    "messages": [
      { "role": "user", "content": "What does a KV cache do?" }
    ]
  }'
POST /v1/chat/completionsapi.weir.example

Dedicated

Need reserved capacity?

DeepSeek R1 0528 runs on the shared serverless pool. Dedicated replicas for it are set up on request, with the same API and the same model ID.

Talk to an engineer

Compare

Similar models