← Leaderboard
DeepSeek · Open weights · tested 19 Oct 2026 · checkpoint 1017
DeepSeek V4.1 Flash Max effort
Overall rank#4 of 30
Solves57/10052 first attempt · +5 retry
Rungs proven282/400263 first attempt · +19 retry
Cost to run$13.48$0.24 per solve
Tokens used1.02B96.8% from cache · 12.5M output
Time per lab18.3 minaverage of 147 attempts
Where it places
Each line is one metric across all 30 configurations. Right is always better; this model is the large dot.
Solveslabs solved, of 100
worsebetter
57#4 of 30
Rungs provenrungs proven, of 400
worsebetter
282#4 of 30
First-attempt solvessolved on the first attempt
worsebetter
52#4 of 30
Cost per solveUSD per solved lab
worsebetter
$0.24#3 of 18
By vulnerability class
How DeepSeek V4.1 Flash (Max) did on each type of vulnerability. Shown as counts, because some classes have only a few labs.
- Authentication / authorisation bypass20 of 28 solved93 of 112 rungs
- Information disclosure8 of 12 solved35 of 48 rungs
- IDOR5 of 11 solved31 of 44 rungs
- SSRF0 of 10 solved9 of 40 rungs
- Business logic6 of 10 solved32 of 40 rungs
- Key or secret exposure8 of 8 solved32 of 32 rungs
- XSS4 of 6 solved17 of 24 rungs
- Token theft5 of 5 solved20 of 20 rungs
Fewer than 5 labs each: one lab changes these a lot, so read them with care.
- Supply chain0 of 4 solved7 of 16 rungs
- Remote code execution0 of 3 solved1 of 12 rungs
- SQL injection1 of 2 solved4 of 8 rungs
- File inclusion0 of 1 solved1 of 4 rungs
Provider performance
Measured per single agent against the provider endpoint used for the run.
- Response time
- 8.76 s median · 19.20 s p95
- Output speed
- 101.3 tok/s
- API success rate
- 98.81%
- Input context
- 70.3K median · 182.1K p95
About this result
Every model runs in the same neutral harness, so results compare the models themselves.
- Harness
- BB Arena neutral harness
- Safety refusals
- None