Skip to content
Get an API key

Live · serverless · 2 regions

Qwen3 Coder 480B A35B

by Alibaba

A very large coding model built for agents that edit many files, run tools and read long repositories.

  • code
  • FP8
Serverless priceUSD
Input
$0.95per 1M tokens
Output
$1.90per 1M tokens

Billed per token, no minimums.

Reference

Specifications

Model ID
qwen3-coder-480b-a35b
Endpoint
POST /v1/chat/completions
Publisher
Alibaba
Tasks
code
Served context
256K tokens
Quantization
FP8
Decode speed
70 tok/smedian, one request
p50 first token
290 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

290 msmedian

p95 312 ms

Decode speed · tok/s

70 tok/smedian

p5 66 tok/s

Show the data
Median of each 15-minute step over the last 24 hours, newest first
StepTime to first tokenDecode speed
15m ago312 ms66 tok/s
30m ago299 ms70 tok/s
45m ago285 ms70 tok/s
1h ago314 ms67 tok/s
1h 15m ago309 ms69 tok/s
1h 30m ago311 ms66 tok/s
1h 45m ago291 ms70 tok/s
2h ago293 ms70 tok/s
2h 15m ago302 ms66 tok/s
2h 30m ago296 ms67 tok/s
2h 45m ago421 ms55 tok/s
3h ago293 ms69 tok/s
3h 15m ago293 ms69 tok/s
3h 30m ago299 ms68 tok/s
3h 45m ago290 ms72 tok/s
4h ago298 ms69 tok/s
4h 15m ago421 ms55 tok/s
4h 30m ago303 ms69 tok/s
4h 45m ago291 ms70 tok/s
5h ago302 ms69 tok/s
5h 15m ago287 ms70 tok/s
5h 30m ago285 ms71 tok/s
5h 45m ago309 ms67 tok/s
6h ago292 ms71 tok/s
6h 15m ago301 ms67 tok/s
6h 30m ago308 ms70 tok/s
6h 45m ago298 ms67 tok/s
7h ago303 ms68 tok/s
7h 15m ago293 ms72 tok/s
7h 30m ago282 ms72 tok/s
7h 45m ago297 ms68 tok/s
8h ago304 ms67 tok/s
8h 15m ago300 ms67 tok/s
8h 30m ago285 ms69 tok/s
8h 45m ago295 ms68 tok/s
9h ago302 ms70 tok/s
9h 15m ago284 ms71 tok/s
9h 30m ago299 ms71 tok/s
9h 45m ago285 ms71 tok/s
10h ago298 ms67 tok/s
10h 15m ago276 ms72 tok/s
10h 30m ago290 ms69 tok/s
10h 45m ago297 ms71 tok/s
11h ago271 ms71 tok/s
11h 15m ago278 ms69 tok/s
11h 30m ago289 ms70 tok/s
11h 45m ago287 ms69 tok/s
12h ago270 ms74 tok/s
12h 15m ago291 ms71 tok/s
12h 30m ago288 ms68 tok/s
12h 45m ago277 ms73 tok/s
13h ago271 ms74 tok/s
13h 15m ago281 ms71 tok/s
13h 30m ago273 ms73 tok/s
13h 45m ago293 ms70 tok/s
14h ago287 ms69 tok/s
14h 15m ago278 ms73 tok/s
14h 30m ago267 ms71 tok/s
14h 45m ago274 ms73 tok/s
15h ago282 ms73 tok/s
15h 15m ago288 ms68 tok/s
15h 30m ago285 ms70 tok/s
15h 45m ago286 ms70 tok/s
16h ago292 ms71 tok/s
16h 15m ago270 ms73 tok/s
16h 30m ago292 ms71 tok/s
16h 45m ago277 ms71 tok/s
17h ago268 ms74 tok/s
17h 15m ago275 ms71 tok/s
17h 30m ago270 ms70 tok/s
17h 45m ago337 ms65 tok/s
18h ago286 ms72 tok/s
18h 15m ago277 ms70 tok/s
18h 30m ago269 ms71 tok/s
18h 45m ago278 ms69 tok/s
19h ago298 ms68 tok/s
19h 15m ago279 ms72 tok/s
19h 30m ago277 ms72 tok/s
19h 45m ago290 ms68 tok/s
20h ago286 ms71 tok/s
20h 15m ago277 ms70 tok/s
20h 30m ago278 ms71 tok/s
20h 45m ago281 ms69 tok/s
21h ago296 ms71 tok/s
21h 15m ago280 ms70 tok/s
21h 30m ago291 ms68 tok/s
21h 45m ago287 ms70 tok/s
22h ago289 ms72 tok/s
22h 15m ago299 ms69 tok/s
22h 30m ago289 ms69 tok/s
22h 45m ago283 ms70 tok/s
23h ago303 ms70 tok/s
23h 15m ago298 ms69 tok/s
23h 30m ago290 ms71 tok/s
23h 45m ago312 ms68 tok/s
24h ago306 ms69 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-coder-480b-a35b",
    "messages": [
      { "role": "user", "content": "What does a KV cache do?" }
    ]
  }'
POST /v1/chat/completionsapi.weir.example

Dedicated

Need reserved capacity?

Qwen3 Coder 480B A35B 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