Mistral-Medium-3.5-128B

Mistral AI
Arena score
1421
Among open models
#32of 148
Among all on Arena
#114
Votes
11,647

Measured on Mistral Medium 3.5, the served version, not a local build. LMArena, CC BY 4.0, published Nov 11, 2026.

Score by category

Top 10 among open models in#7German

CategoryScore
  • Coding1461
  • German1460
  • Software & IT1458
  • French1448
  • Hard prompts, English1444
  • Chinese1444
  • Applied math1440
  • Legal & government1437
  • Hard prompts1436
  • English1433
  • Expert1431
  • Math1430
  • Polish1429
  • Science1428
  • Business & finance1426
  • Multi-turn1424
  • Instructions1416
  • Long prompts1415
  • Spanish1412
  • Excluding ties1410
  • Medicine & health1407
  • Non-English1405
  • Russian1397
  • Korean1383
  • Writing & literature1383
  • Media & sports1383
  • Creative writing1374

Score by size

Plot against
Best for its size

About as strong, smaller

Compare

Discussion

Built on it

Quantizations
43
Fine-tunes
11
Adapters
1
Merges
1

Ways to run it

Most downloaded

BenchClub runs

Not runnable yet

Runs open once a Mistral-Medium-3.5-128B build is registered.

More from Mistral AI

See more

Other 80–300B models

See more