Gemini 3.5 Flash

Gemini 3.5 Flash is a model from Google, released May 19, 2026. Its best result is #6 of 107 on MedCode.

Release Date: May 19, 2026

Developer Google ย ๐Ÿ‡บ๐Ÿ‡ธ
Context Window 1M
Max Output Tokens 66k
Token Costs (in/out) $1.50/9.00
Weights Private
Input Modalities

Accuracy

56.16 %

Avg. Cost (In/Out)

$ 1.50 / $ 9.00

Latency

10 min 27 s

Vals Index
BenchmarksAccuracyRankings

26.75%

ยฑ4.10
46/79

63.55%

ยฑ2.75
27/76

57.86%

ยฑ0.23
14/81

62.96%

ยฑ1.66
26/28

30.77%

ยฑ3.21
60/98

55.83%

ยฑ2.11
6/107

76.57%

ยฑ1.92
75/109

68.12%

ยฑ0.91
19/98

31.00%

ยฑ4.65
37/53

59.47%

ยฑ1.28
33/50

21.65%

ยฑ2.96
58/89

74.37%

ยฑ0.85
37/145

48.68%

ยฑ4.73
76/132

92.68%

ยฑ1.46
14/138

87.60%

ยฑ0.95
12/143

83.60%

ยฑ0.85
52/151

89.52%

ยฑ0.31
10/138

88.27%

ยฑ0.77
8/93

0.00%

ยฑ0.00
48/66

52.74%

ยฑ4.47
18/35

78.80%

ยฑ1.83
28/88

6.06%

ยฑ1.51
57/68

5.71%

ยฑ2.79
25/47
Proprietary BenchmarksAcademic BenchmarksIndustry Partners
Vals
Default Provider : Google
Temperature: Default
Top P: Default
Top K: Default
Max Output Tokens: 65,536
Reasoning Effort: high

Compared with

Updates

May 19, 2026

The model has a 1M-token context window. Evaluations were run using a reasoning effort of โ€œhighโ€, a temperature of 1.0, and max output tokens set to 65k.

Congrats to the Google team on the strong release!