Skip to content
Get an API key

Live · serverless · 3 regions

Qwen3 8B

by Alibaba

An 8B model with solid multilingual coverage, priced for high-volume routing and tagging jobs.

  • chat
  • BF16
Serverless priceUSD
Input
$0.07per 1M tokens
Output
$0.12per 1M tokens

Billed per token, no minimums.

Reference

Specifications

Model ID
qwen3-8b
Endpoint
POST /v1/chat/completions
Publisher
Alibaba
Tasks
chat
Served context
128K tokens
Quantization
BF16
Decode speed
270 tok/smedian, one request
p50 first token
68 mslast 24 hours
Regions
us-east · us-west · eu-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

68 msmedian

p95 89 ms

Decode speed · tok/s

270 tok/smedian

p5 224 tok/s

Show the data
Median of each 15-minute step over the last 24 hours, newest first
StepTime to first tokenDecode speed
15m ago72 ms270 tok/s
30m ago71 ms265 tok/s
45m ago68 ms270 tok/s
1h ago73 ms262 tok/s
1h 15m ago67 ms270 tok/s
1h 30m ago71 ms269 tok/s
1h 45m ago69 ms277 tok/s
2h ago69 ms264 tok/s
2h 15m ago69 ms270 tok/s
2h 30m ago73 ms254 tok/s
2h 45m ago72 ms267 tok/s
3h ago72 ms265 tok/s
3h 15m ago74 ms256 tok/s
3h 30m ago67 ms268 tok/s
3h 45m ago69 ms276 tok/s
4h ago72 ms257 tok/s
4h 15m ago68 ms266 tok/s
4h 30m ago70 ms262 tok/s
4h 45m ago73 ms264 tok/s
5h ago74 ms263 tok/s
5h 15m ago70 ms263 tok/s
5h 30m ago70 ms268 tok/s
5h 45m ago71 ms273 tok/s
6h ago71 ms273 tok/s
6h 15m ago68 ms267 tok/s
6h 30m ago67 ms269 tok/s
6h 45m ago92 ms220 tok/s
7h ago69 ms273 tok/s
7h 15m ago71 ms271 tok/s
7h 30m ago66 ms281 tok/s
7h 45m ago67 ms268 tok/s
8h ago66 ms277 tok/s
8h 15m ago67 ms280 tok/s
8h 30m ago69 ms266 tok/s
8h 45m ago70 ms264 tok/s
9h ago95 ms212 tok/s
9h 15m ago64 ms281 tok/s
9h 30m ago67 ms275 tok/s
9h 45m ago64 ms285 tok/s
10h ago65 ms282 tok/s
10h 15m ago67 ms281 tok/s
10h 30m ago68 ms270 tok/s
10h 45m ago64 ms276 tok/s
11h ago89 ms233 tok/s
11h 15m ago66 ms269 tok/s
11h 30m ago66 ms280 tok/s
11h 45m ago69 ms276 tok/s
12h ago94 ms213 tok/s
12h 15m ago66 ms273 tok/s
12h 30m ago93 ms224 tok/s
12h 45m ago67 ms266 tok/s
13h ago67 ms281 tok/s
13h 15m ago65 ms275 tok/s
13h 30m ago67 ms275 tok/s
13h 45m ago69 ms265 tok/s
14h ago66 ms272 tok/s
14h 15m ago64 ms272 tok/s
14h 30m ago66 ms273 tok/s
14h 45m ago62 ms284 tok/s
15h ago63 ms275 tok/s
15h 15m ago64 ms278 tok/s
15h 30m ago68 ms267 tok/s
15h 45m ago63 ms288 tok/s
16h ago63 ms274 tok/s
16h 15m ago68 ms267 tok/s
16h 30m ago67 ms276 tok/s
16h 45m ago63 ms277 tok/s
17h ago66 ms268 tok/s
17h 15m ago62 ms275 tok/s
17h 30m ago66 ms276 tok/s
17h 45m ago66 ms269 tok/s
18h ago64 ms273 tok/s
18h 15m ago64 ms286 tok/s
18h 30m ago69 ms276 tok/s
18h 45m ago65 ms276 tok/s
19h ago69 ms264 tok/s
19h 15m ago65 ms270 tok/s
19h 30m ago68 ms273 tok/s
19h 45m ago69 ms263 tok/s
20h ago66 ms268 tok/s
20h 15m ago66 ms270 tok/s
20h 30m ago68 ms262 tok/s
20h 45m ago67 ms274 tok/s
21h ago67 ms279 tok/s
21h 15m ago71 ms263 tok/s
21h 30m ago71 ms262 tok/s
21h 45m ago72 ms270 tok/s
22h ago71 ms272 tok/s
22h 15m ago71 ms261 tok/s
22h 30m ago68 ms267 tok/s
22h 45m ago89 ms224 tok/s
23h ago71 ms268 tok/s
23h 15m ago70 ms271 tok/s
23h 30m ago70 ms260 tok/s
23h 45m ago68 ms276 tok/s
24h ago72 ms258 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": "qwen3-8b",
    "messages": [
      { "role": "user", "content": "What does a KV cache do?" }
    ]
  }'
POST /v1/chat/completionsapi.weir.example

Dedicated

Need reserved capacity?

Qwen3 8B 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