[mle]/DaBench

DeepSeek V4.1 Flash 1017 vs Mistral Large 4

Side A
Model page
Side B
Model page
Overall
56
vs
Overall
28

DeepSeek V4.1 Flash 1017 is clearly better. It wins 62 of the 74 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. Both models spend about the same on thinking per test case.

Thinking tokensAnswer tokens
TaskAB
Overall
5626.0s
5.3K, 91% thinking6.2K, 72% thinking
281m 56s
Agents
5826.7s
3.3K, 71% thinking795, no thinking
1126.1s
Structured Extraction
8117.2s
4.2K, 93% thinking288, no thinking
153.78s
Coding
4840.8s
9.4K, 95% thinking9.3K, 37% thinking
332m 48s
Grounded Answers
7119.8s
4.4K, 91% thinking8.1K, 89% thinking
382m 45s
Instruction Following
7625.6s
5.5K, 91% thinking12.9K, 84% thinking
693m 47s
Business Logic
6726.6s
6.6K, 97% thinking8.8K, 77% thinking
242m 40s
Security & permissions
4826.6s
4.7K, 87% thinking3.7K, 83% thinking
331m 9s
Diagnosis & investigation
020.2s
3.2K, 74% thinking4.4K, 85% thinking
01m 19s

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)
56 tok/s
Machine
Hosted
Hosted
Memory at 32K
—
—
Context
1M (better)
512K
Cost per run
$0.75 (better)
$2.67
Cost per case
$0.0049 (better)
$0.017
Time per case
26.0s (better)
1m 56s
Quantization
—
—
Harness
openrouter
openrouter

The race

No lead changes

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

DeepSeek V4.1 Flash 1017Mistral Large 4
025507510050100150Perfect run5628

Every test case

153 test cases

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

62 won by DeepSeek V4.1 Flash 101712 won by Mistral Large 479 tied