Mistral Medium 3.5
mistral-mistral-medium-3-5
fields
| nameREQ | Mistral Medium 3.5 | mistral.ai | 2026-04-30 |
| vendorREQ | Mistral AI | mistral.ai | 2026-04-30 |
| released_at | 2026-04-29 | mistral.ai | 2026-04-30 |
| modality | text, code, vision | mistral.ai | 2026-04-30 |
| access | open-weights | mistral.ai | 2026-04-30 |
| context_window | 262144 | mistral.ai | 2026-04-30 |
| parameters | 128B | mistral.ai | 2026-04-30 |
| benchmark_score | 77.6 | mistral.ai | 2026-04-30 |
| benchmark_name | SWE-Bench Verified | mistral.ai | 2026-04-30 |
| announcement_url | https://mistral.ai/news/vibe-remote-agents-mistral-medium-3-5 | mistral.ai | 2026-04-30 |
history · 10 fields · 10 revisions
name1 revision
Mistral Medium 3.5
current
Mistral Medium 3.5: a dense 128B model with a 256k context window. Scores 77.6% on SWE-Bench Verified. Released as open weights, under a modified MIT license.
vendor1 revision
Mistral AI
current
Mistral Medium 3.5: a dense 128B model with a 256k context window. Scores 77.6% on SWE-Bench Verified. Released as open weights, under a modified MIT license.
released_at1 revision
2026-04-29
current
Mistral Medium 3.5: a dense 128B model with a 256k context window. Scores 77.6% on SWE-Bench Verified. Released as open weights, under a modified MIT license.
modality1 revision
text, code, vision
current
Mistral Medium 3.5: a dense 128B model with a 256k context window. Scores 77.6% on SWE-Bench Verified. Released as open weights, under a modified MIT license.
access1 revision
open-weights
current
Mistral Medium 3.5: a dense 128B model with a 256k context window. Scores 77.6% on SWE-Bench Verified. Released as open weights, under a modified MIT license.
context_window1 revision
262144
current
Mistral Medium 3.5: a dense 128B model with a 256k context window. Scores 77.6% on SWE-Bench Verified. Released as open weights, under a modified MIT license.
parameters1 revision
128B
current
Mistral Medium 3.5: a dense 128B model with a 256k context window. Scores 77.6% on SWE-Bench Verified. Released as open weights, under a modified MIT license.
benchmark_score1 revision
77.6
current
Mistral Medium 3.5: a dense 128B model with a 256k context window. Scores 77.6% on SWE-Bench Verified. Released as open weights, under a modified MIT license.
benchmark_name1 revision
SWE-Bench Verified
current
Mistral Medium 3.5: a dense 128B model with a 256k context window. Scores 77.6% on SWE-Bench Verified. Released as open weights, under a modified MIT license.
announcement_url1 revision
https://mistral.ai/news/vibe-remote-agents-mistral-medium-3-5
current
Mistral Medium 3.5: a dense 128B model with a 256k context window. Scores 77.6% on SWE-Bench Verified. Released as open weights, under a modified MIT license.