[mle]/DaBench

DeepSeek V4.1 Flash 1017 vs Qwen3.8 Flash

Side A
Model page
Side B
Model page
Overall
56
vs
Overall
32

DeepSeek V4.1 Flash 1017 is clearly better. It wins 69 of the 88 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. DeepSeek V4.1 Flash 1017 spends 1.5× as many thinking tokens per test case.

Thinking tokensAnswer tokens
TaskAB
Overall
5626.0s
5.3K, 91% thinking4K, 81% thinking
3258.5s
Agents
5826.7s
3.3K, 71% thinking4.2K, 76% thinking
491m 25s
Structured Extraction
8117.2s
4.2K, 93% thinking3K, 90% thinking
3336.9s
Coding
4840.8s
9.4K, 95% thinking5.8K, 61% thinking
471m 21s
Grounded Answers
7119.8s
4.4K, 91% thinking3.1K, 90% thinking
2641.4s
Instruction Following
7625.6s
5.5K, 91% thinking5K, 89% thinking
371m 4s
Business Logic
6726.6s
6.6K, 97% thinking3.9K, 91% thinking
1746.8s
Security & permissions
4826.6s
4.7K, 87% thinking3.2K, 83% thinking
5054.1s
Diagnosis & investigation
020.2s
3.2K, 74% 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
243 tok/s (better)
77 tok/s
Machine
Hosted
Hosted
Memory at 32K
—
—
Context
1M
1M
Cost per run
$0.75
$0.46 (better)
Cost per case
$0.0049
$0.003 (better)
Time per case
26.0s (better)
58.5s
Quantization
—
—
Harness
openrouter
openrouter

The race

3 lead changes

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

DeepSeek V4.1 Flash 1017Qwen3.8 Flash
025507510050100150Perfect run5632

Every test case

153 test cases

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

69 won by DeepSeek V4.1 Flash 101719 won by Qwen3.8 Flash65 tied