Ling 3.1 Flash vs Mellum2.1 Thinking, head to head

Side A

Side B

Overall

Ling 3.1 Flash is clearly better. It wins 67 of the 71 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

TaskALing 3.1 FlashBMellum2.1 ThinkingLing 3.1 FlashMellum2.1 Thinking

Overall

451m 16s

7.9K10.3K

91m 56s

Agents

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

4155.9s

4.5K10.8K

92m 3s

Structured Extraction

Turns logs, diffs and records into exact JSON

711m 11s

8.1K8.3K

51m 36s

Coding

Writes code and SQL that pass hidden tests

241m 42s

11.3K11.3K

52m 11s

Grounded Answers

Answers from the given documents with sources, never guesses

5258.9s

6.7K7.3K

131m 24s

Instruction Following

Applies every rule, including ones changed mid-conversation

761m 55s

11.4K17.3K

283m 10s

Business Logic

Computes billing, scheduling and pricing rules exactly

381m 22s

9.7K9K

01m 39s

Security & permissions

Ignores planted instructions, still does the allowed work

5456.1s

5.2K9.1K

81m 40s

Diagnosis & investigation

Chooses diagnostic checks, repairs supported causes and verifies recovery

042.1s

3.2K6.8K

01m 15s

What you'd run

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

SpecALing 3.1 FlashBMellum2.1 Thinking

Speed

112 tok/s

90 tok/s

Machine

Hosted

Apple M5 Pro, 48 GB

Memory at 32K

—

12.9 GB

Context

256K

128K

Cost per run

Not measured

$2.46 est.

Cost per case

Not measured

$0.016 est.

Time per case

1m 16s

1m 56s

Quantization

—

Q8_0

Harness

openrouter

llama.cpp

The race

No lead changes

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

Ling 3.1 FlashMellum2.1 Thinking

Every test case

153 test cases

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

Agents81

Structured Extraction151

Coding41

Grounded Answers90

Instruction Following120

Business Logic80

Security & permissions111

Diagnosis & investigation00

67 won by Ling 3.1 Flash4 won by Mellum2.1 Thinking82 tied