Apodex 1.1 Mini (free) vs Claude Haiku 5.5, head to head

Side A

Side B

Overall

Claude Haiku 5.5 is clearly better. It wins 82 of the 84 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. Apodex 1.1 Mini (free) spends 2.1× as many thinking tokens per test case.

Thinking tokensAnswer tokens

TaskAApodex 1.1 Mini (free)BClaude Haiku 5.5Apodex 1.1 Mini (free)Claude Haiku 5.5

Overall

2535.1s

8.5K3.9K

7025.5s

Agents

Runs multi-step jobs through tools, within a written policy

2639.4s

6.2K3.7K

6139.5s

Structured Extraction

Turns logs, diffs and records into exact JSON

5227.2s

7.7K2.6K

9013.5s

Coding

Writes code and SQL that pass hidden tests

446.5s

12.1K6.8K

8630.6s

Grounded Answers

Answers from the given documents with sources, never guesses

2926.0s

6.9K2.2K

7613.9s

Instruction Following

Applies every rule, including ones changed mid-conversation

3644.5s

11.4K4.3K

9431.5s

Business Logic

Computes billing, scheduling and pricing rules exactly

1435.4s

10K5.5K

6726.0s

Security & permissions

Ignores planted instructions, still does the allowed work

4027.3s

5.8K2.7K

6723.4s

Diagnosis & investigation

Chooses diagnostic checks, repairs supported causes and verifies recovery

034.6s

5.1K2.3K

1725.5s

What you'd run

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

SpecAApodex 1.1 Mini (free)BClaude Haiku 5.5

Speed

278 tok/s

225 tok/s

Machine

Hosted

Memory at 32K

—

Context

256K

Cost per run

Not measured

$0.42

Cost per case

Not measured

$0.0028

Time per case

35.1s

25.5s

Quantization

BF16

—

Harness

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)Claude Haiku 5.5

Every test case

153 test cases

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

Agents010

Structured Extraction19

Coding018

Grounded Answers010

Instruction Following015

Business Logic011

Security & permissions18

Diagnosis & investigation01

2 won by Apodex 1.1 Mini (free)82 won by Claude Haiku 5.569 tied