Sign in

Qwen3.5-27B

Qwen

Hugging FaceArtificial Analysis

BenchClub cloud

Pass rate
93%
25 of 27 checks
Speed
35.5 tok/s
decode, short prompt
At 8k context
27.7 tok/s
decode
Cloud runs
3
speeds within 2.5%
  • Coding8/989%
  • Tool use9/9100%
  • Long context8/989%
Suite
[email protected]
Machine
NVIDIA L40S · 48 GB
Engine
ollama 0.40.14

Headline numbers come only from cloud runs. Runs on members’ own machines are listed below and don’t change them.

Arena score
1409
Among open models
#40of 144
Among all on Arena
#143
Votes
28,647

Measured on Qwen3.5-27B, the served version, not a local build. LMArena, CC BY 4.0, published Oct 28, 2026.

Score by category

CategoryScore
  • Chinese1478
  • Science1436
  • Software & IT1431
  • Applied math1429
  • Math1429
  • Hard prompts, English1428
  • Expert1427
  • Coding1426
  • Medicine & health1425
  • English1425
  • Hard prompts1414
  • Long prompts1413
  • Multi-turn1410
  • French1410
  • Spanish1407
  • Legal & government1406
  • Business & finance1399
  • German1393
  • Instructions1393
  • Excluding ties1392
  • Non-English1390
  • Russian1390
  • Writing & literature1381
  • Polish1375
  • Media & sports1363
  • Creative writing1362
  • Korean1359
  • Japanese1346

BenchClub runs

Run this model
Run by

PassedFailedNot supported