[mle]/DaBench

Ling 3.1 Flash vs Qwen3.8 Flash

Side A
Model page
Side B
Model page
Overall
45
vs
Overall
32

Ling 3.1 Flash is clearly better. It wins 50 of the 71 test cases where they differ.

Learn to build evals like this in the AI Engineering Academy

Task by task

A score turns orange only when that model wins clearly more of the task's test cases. Under each score, the average time per test case on the machine it ran on. Ling 3.1 Flash spends 1.8× as many thinking tokens per test case.

Thinking tokensAnswer tokens
TaskAB
Overall
451m 16s
7.9K, 73% thinking4K, 81% thinking
3258.5s
Agents
4155.9s
4.5K, 66% thinking4.2K, 76% thinking
491m 25s
Structured Extraction
711m 11s
8.1K, 69% thinking3K, 90% thinking
3336.9s
Coding
241m 42s
11.3K, 89% thinking5.8K, 61% thinking
471m 21s
Grounded Answers
5258.9s
6.7K, 74% thinking3.1K, 90% thinking
2641.4s
Instruction Following
761m 55s
11.4K, 77% thinking5K, 89% thinking
371m 4s
Business Logic
381m 22s
9.7K, 56% thinking3.9K, 91% thinking
1746.8s
Security & permissions
5456.1s
5.2K, 74% thinking3.2K, 83% thinking
5054.1s
Diagnosis & investigation
042.1s
3.2K, 72% thinking3.3K, 73% thinking
058.4s

What you'd run

The build behind each score. Orange marks the faster, smaller, cheaper or longer-context side.

SpecAB
Speed
112 tok/s (better)
77 tok/s
Machine
Hosted
Hosted
Memory at 32K
—
—
Context
256K
1M (better)
Cost per run
Not measured
$0.46
Cost per case
Not measured
$0.003
Time per case
1m 16s
58.5s (better)
Quantization
—
—
Harness
openrouter
openrouter

The race

5 lead changes

Each model's running score as the test cases go by. The dashed line is a perfect run.

Ling 3.1 FlashQwen3.8 Flash
025507510050100150Perfect run4532

Every test case

153 test cases

Each square is one test case, colored by the model that scored higher.

50 won by Ling 3.1 Flash21 won by Qwen3.8 Flash82 tied