Qwen3.5-9B
Qwen Team (Alibaba Cloud)
Parameters
—
Architecture
—
Released
—
License
—
About
No detailed description available for this model.
Benchmark Scores
| Benchmark | Score | Date | Standing |
|---|---|---|---|
|
Terminal-Bench 2.1 (Terminus-2)
coding_agent
|
0.44%
|
25.06.2026 |
#43/44
better than 2%
|
|
Terminal-Bench 2.1 (Claude Code)
coding_agent
|
18.90
|
25.06.2026 |
#8/8
better than 0%
|
|
SWE-bench Verified
coding_agent
|
60.44%
|
25.06.2026 |
#45/60
better than 25%
|
|
NL2Repo
coding_agent
|
9.08%
|
25.06.2026 |
#37/41
better than 10%
|
|
Claw-Eval Avg
coding_agent
|
63.36%
|
25.06.2026 |
#30/40
better than 26%
|
|
SWE Atlas - QnA
coding_agent
|
9.20
|
25.06.2026 |
#24/24
better than 0%
|
|
SWE Atlas - RF
coding_agent
|
4.30
|
25.06.2026 |
#14/14
better than 0%
|
|
SWE Atlas - TW
coding_agent
|
4.40
|
25.06.2026 |
#20/20
better than 0%
|
|
VITA-Bench
general_agent
|
29.02%
|
— |
#10/16
better than 40%
|
|
BFCL-V4
general_agent
|
85.07%
|
— |
#8/27
better than 73%
|
|
TAU2-Bench
general_agent
|
77.78%
|
— |
#6/21
better than 75%
|
|
TAU3-Bench
general_agent
|
11.96%
|
— |
#14/21
better than 35%
|
|
MCPMark
general_agent
|
12.04%
|
— |
#13/15
better than 14%
|
|
Workspace Bench
|
75.74%
|
— |
#2/5
better than 75%
|
|
BrowseComp
general_agent
|
5.84%
|
— |
#41/45
better than 7%
|
|
SciCode
|
50.86%
|
— |
#16/27
better than 42%
|
|
Gaokao 2026
|
100.00%
|
— |
#1/8
better than 100%
|
|
AIME 26
stem_reasoning
|
85.97%
|
— |
#33/45
better than 27%
|
|
HMMT Feb 26
stem_reasoning
|
65.16%
|
— |
#33/44
better than 26%
|
|
IMOAnswerBench
stem_reasoning
|
67.05%
|
— |
#24/30
better than 21%
|
|
IFBench
instruction_following
|
67.97%
|
— |
#31/42
better than 27%
|
|
Humanity's Last Exam
stem_reasoning
|
23.11%
|
— |
#55/71
better than 23%
|
|
GPQA
reasoning
|
100.00%
|
— |
#1/14
better than 92%
|
|
SWE-bench Multilingual
coding_agent
|
45.33%
|
25.06.2026 |
#42/48
better than 13%
|
|
MCP-Atlas
general_agent
|
45.43%
|
— |
#47/54
better than 13%
|
|
IFEval
instruction_following
|
91.53%
|
— |
#11/26
better than 60%
|
|
SWE-bench Pro
coding_agent
|
37.68%
|
25.06.2026 |
#64/77
better than 17%
|
|
AA-LCR
long_context
|
75.00%
|
— |
#25/42
better than 41%
|
What is a good Terminal-Bench 2.1 (Terminus-2) score? What is a good Terminal-Bench 2.1 (Claude Code) score? What is a good SWE-bench Verified score?
No comments yet
Be the first to share your thoughts!