[mle]/DaBench

DeepSeek V4.1 Flash 1017 vs Ling 3.1 Flash

Side A
Model page
Side B
Model page
Overall
56
vs
Overall
45

DeepSeek V4.1 Flash 1017 is clearly better. It wins 39 of the 56 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 1.2× as many thinking tokens per test case.

Thinking tokensAnswer tokens
TaskAB
Overall
5626.0s
5.3K, 91% thinking7.9K, 73% thinking
451m 16s
Agents
5826.7s
3.3K, 71% thinking4.5K, 66% thinking
4155.9s
Structured Extraction
8117.2s
4.2K, 93% thinking8.1K, 69% thinking
711m 11s
Coding
4840.8s
9.4K, 95% thinking11.3K, 89% thinking
241m 42s
Grounded Answers
7119.8s
4.4K, 91% thinking6.7K, 74% thinking
5258.9s
Instruction Following
7625.6s
5.5K, 91% thinking11.4K, 77% thinking
761m 55s
Business Logic
6726.6s
6.6K, 97% thinking9.7K, 56% thinking
381m 22s
Security & permissions
4826.6s
4.7K, 87% thinking5.2K, 74% thinking
5456.1s
Diagnosis & investigation
020.2s
3.2K, 74% thinking3.2K, 72% thinking
042.1s

What you'd run

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

SpecAB
Speed
243 tok/s (better)
112 tok/s
Machine
Hosted
Hosted
Memory at 32K
—
—
Context
1M (better)
256K
Cost per run
$0.75
Not measured
Cost per case
$0.0049
Not measured
Time per case
26.0s (better)
1m 16s
Quantization
—
—
Harness
openrouter
openrouter

The race

2 lead changes

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

DeepSeek V4.1 Flash 1017Ling 3.1 Flash
025507510050100150Perfect run5645

Every test case

153 test cases

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

39 won by DeepSeek V4.1 Flash 101717 won by Ling 3.1 Flash97 tied