Mellum2.1 Thinking vs Muse Spark 1.3 Contributor, head to head

Side A

Side B

Overall

Muse Spark 1.3 Contributor is clearly better. It wins 101 of the 103 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. Muse Spark 1.3 Contributor spends 1.2× as many thinking tokens per test case.

Thinking tokensAnswer tokens

TaskAMellum2.1 ThinkingBMuse Spark 1.3 ContributorMellum2.1 ThinkingMuse Spark 1.3 Contributor

Overall

91m 56s

10.3K8.7K

661m 42s

Agents

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

92m 3s

10.8K14.7K

573m 16s

Structured Extraction

Turns logs, diffs and records into exact JSON

51m 36s

8.3K4.2K

9037.9s

Coding

Writes code and SQL that pass hidden tests

52m 11s

11.3K10.8K

482m 1s

Grounded Answers

Answers from the given documents with sources, never guesses

131m 24s

7.3K6.1K

811m 1s

Instruction Following

Applies every rule, including ones changed mid-conversation

283m 10s

17.3K8.7K

841m 18s

Business Logic

Computes billing, scheduling and pricing rules exactly

01m 39s

9K6K

761m 7s

Security & permissions

Ignores planted instructions, still does the allowed work

81m 40s

9.1K8.2K

761m 59s

Diagnosis & investigation

Chooses diagnostic checks, repairs supported causes and verifies recovery

01m 15s

6.8K17.2K

173m 34s

What you'd run

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

SpecAMellum2.1 ThinkingBMuse Spark 1.3 Contributor

Speed

90 tok/s

107 tok/s

Machine

Apple M5 Pro, 48 GB

Hosted

Memory at 32K

12.9 GB

—

Context

128K

Cost per run

$2.46 est.

$0.41

Cost per case

$0.016 est.

$0.0027

Time per case

1m 56s

1m 42s

Quantization

Q8_0

—

Harness

llama.cpp

openrouter

The race

No lead changes

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

Mellum2.1 ThinkingMuse Spark 1.3 Contributor

Every test case

153 test cases

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

Agents112

Structured Extraction018

Coding19

Grounded Answers015

Instruction Following014

Business Logic016

Security & permissions016

Diagnosis & investigation01

2 won by Mellum2.1 Thinking101 won by Muse Spark 1.3 Contributor50 tied