LFM2.5 24B A2B vs Spark X2.6 4B, head to head

Side A

Side B

Overall

Spark X2.6 4B is clearly better. It wins 22 of the 29 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. LFM2.5 24B A2B spends 1.2× as many thinking tokens per test case.

Thinking tokensAnswer tokens

TaskALFM2.5 24B A2BBSpark X2.6 4BLFM2.5 24B A2BSpark X2.6 4B

Overall

71m 53s

9.2K8.2K

151m 3s

Agents

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

123m 49s

18.5K11.5K

141m 32s

Structured Extraction

Turns logs, diffs and records into exact JSON

1054.9s

4.4K5.3K

1039.5s

Coding

Writes code and SQL that pass hidden tests

11m 15s

6.1K9.1K

41m 8s

Grounded Answers

Answers from the given documents with sources, never guesses

557.8s

4.7K6.6K

1050.2s

Instruction Following

Applies every rule, including ones changed mid-conversation

162m 44s

13.4K11.9K

371m 29s

Business Logic

Computes billing, scheduling and pricing rules exactly

51m 2s

5K4.9K

1036.3s

Security & permissions

Ignores planted instructions, still does the allowed work

101m 52s

9K7K

3154.3s

Diagnosis & investigation

Chooses diagnostic checks, repairs supported causes and verifies recovery

04m 12s

19.9K11.6K

01m 30s

What you'd run

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

SpecALFM2.5 24B A2BBSpark X2.6 4B

Speed

82 tok/s

136 tok/s

Machine

Apple M5 Pro, 48 GB

RTX 5090, 32 GB

Memory at 32K

25.0 GB

9.4 GB

Context

128K

605K

Cost per run

Not measured

$1.61 est.

Cost per case

Not measured

$0.011 est.

Time per case

1m 53s

1m 3s

Quantization

Q8_0

BF16

Harness

llama.cpp

The race

2 lead changes

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

LFM2.5 24B A2BSpark X2.6 4B

Every test case

153 test cases

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

Agents13

Structured Extraction22

Coding13

Grounded Answers01

Instruction Following27

Business Logic01

Security & permissions15

Diagnosis & investigation00

7 won by LFM2.5 24B A2B22 won by Spark X2.6 4B124 tied