[mle]/DaBench

Apodex 1.1 Mini (free) vs Flint

Side A
Model page
Side B
Model page
Overall
25
vs
Overall
21

Too close to call. Apodex 1.1 Mini (free) wins 20 of the 32 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. Flint spends 1.6× as many thinking tokens per test case.

Thinking tokensAnswer tokens
TaskAB
Overall
2535.1s
8.5K, 75% thinking10.9K, 90% thinking
211m 17s
Agents
2639.4s
6.2K, 69% thinking4.8K, 85% thinking
1447.7s
Structured Extraction
5227.2s
7.7K, 68% thinking10.9K, 87% thinking
291m 17s
Coding
446.5s
12.1K, 94% thinking12K, 98% thinking
141m 19s
Grounded Answers
2926.0s
6.9K, 70% thinking8.7K, 93% thinking
2957.7s
Instruction Following
3644.5s
11.4K, 79% thinking22.2K, 93% thinking
352m 28s
Business Logic
1435.4s
10K, 56% thinking11.7K, 78% thinking
101m 20s
Security & permissions
4027.3s
5.8K, 76% thinking7.1K, 90% thinking
3552.6s
Diagnosis & investigation
034.6s
5.1K, 84% thinking7.3K, 99% thinking
057.0s

What you'd run

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

SpecAB
Speed
278 tok/s (better)
149 tok/s
Machine
Hosted
Hosted
Memory at 32K
—
—
Context
256K
256K
Cost per run
Not measured
Not measured
Cost per case
Not measured
Not measured
Time per case
35.1s (better)
1m 17s
Quantization
BF16
BF16
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.

Apodex 1.1 Mini (free)Flint
0255075100501001502521

Every test case

153 test cases

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

20 won by Apodex 1.1 Mini (free)12 won by Flint121 tied