Mistral Medium 3.5

Mistral Medium 3.5 is an open-weight model from Mistral, released April 29, 2026.

Release Date: Apr 29, 2026

Developer Mistral ย ๐Ÿ‡ซ๐Ÿ‡ท
Context Window 262k
Max Output Tokens 80k
Token Costs (in/out) $1.50/7.50
Weights Open
Input Modalities

Accuracy

34.71 %

Avg. Cost (In/Out)

$ 1.50 / $ 7.50

Latency

24 min 28 s

Vals Index
BenchmarksAccuracyRankings

5.13%

ยฑ1.37
72/79

11.19%

ยฑ1.64
74/76

32.10%

ยฑ0.97
71/81

9.13%

ยฑ2.00
88/95

33.75%

ยฑ1.15
85/107

67.73%

ยฑ2.01
99/109

28.89%

ยฑ0.84
95/98

9.00%

ยฑ2.88
49/53

71.97%

ยฑ2.21
28/29

4.48%

ยฑ1.47
83/86

67.99%

ยฑ0.92
105/145

2.89%

ยฑ1.71
123/129

34.85%

ยฑ2.39
134/138

75.33%

ยฑ0.42
111/138

66.40%

ยฑ2.11
72/88

38.95%

ยฑ2.70
69/75
Proprietary BenchmarksAcademic BenchmarksIndustry Partners
Vals
Default Provider : Mistral
Temperature: 0.7
Top P: 0.95
Top K: Default
Max Output Tokens: 80,000
Reasoning Effort: high

Updates

May 1, 2026

We evaluated Mistral Medium 3.5 on our suite of benchmarks. Here are the key takeaways:

  • Mistralโ€™s new reasoning-mode Medium 3.5 lands at #32 of 46 on the Vals Index (52.77%), and #10 of 18 among open-weight models.
  • It is a sizable jump over the prior Mistral Large 3 across most of the suite: +28 points on Finance Agent (46.1% vs 18.1%), +29 points on the SWE-bench Verified subset of the Vals Index (64.7% vs 35.3%), +21 points on Terminal-Bench 2.0 (30.3% vs 9.0%), and +13 points on SAGE (37.6% vs 24.6%).
  • One regression worth flagging: Case Law (v2) drops 17 points (44.2% vs 61.4%).

All evals were run via the Mistral API at temperature 0.7, top-p 0.95, with reasoning_effort: high and an 80k max output budget.