L
LongCat Flash Lite
Last updated: Recently
Model Type
Text
Max Context
-
Max Output
-
Parameter Scale
-
LongCat Flash Lite by LongCat: Text model; TTFT 4.209s, 129.0 tok/s.
Input Modality
-
Output Modality
-
Inference Speed
128.958 tokens/s
Latest Release Date
1/28/2026
SDK Ecosystem
-
artificial-analysismanufactureraa-bootstrap
Model Overview Fields
name
LongCat Flash Lite
Release date
1/28/2026
Performance
128.958 tokens/s
model_creator.name
LongCat
AA Evaluation Scores
AA Intelligence Index
23.9
AA Coding Index
16.5
AA Math Index
-
MMLU Pro
-
GPQA
0.636
HLE
0.06
LiveCodeBench
-
SciCode
0.284
Math-500
-
AIME
-
AIME 2025
-
IFBench
0.431
LCR
0.257
TerminalBench Hard
0.106
TAU2
0.795
Output Speed (tokens/s)
128.958 tok/s
TTFT (s)
4.209s
First Answer Token (s)
4.209s

