DeepSeek V4.1 Flash 1017 vs Mellum2.1 Thinking, head to head

Side A

Side B

Overall

DeepSeek V4.1 Flash 1017 is clearly better. It wins 86 of the 87 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. Mellum2.1 Thinking spends 1.3× as many thinking tokens per test case.

Thinking tokensAnswer tokens

TaskADeepSeek V4.1 Flash 1017BMellum2.1 ThinkingDeepSeek V4.1 Flash 1017Mellum2.1 Thinking

Overall

5626.0s

5.3K10.3K

91m 56s

Agents

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

5826.7s

3.3K10.8K

92m 3s

Structured Extraction

Turns logs, diffs and records into exact JSON

8117.2s

4.2K8.3K

51m 36s

Coding

Writes code and SQL that pass hidden tests

4840.8s

9.4K11.3K

52m 11s

Grounded Answers

Answers from the given documents with sources, never guesses

7119.8s

4.4K7.3K

131m 24s

Instruction Following

Applies every rule, including ones changed mid-conversation

7625.6s

5.5K17.3K

283m 10s

Business Logic

Computes billing, scheduling and pricing rules exactly

6726.6s

6.6K9K

01m 39s

Security & permissions

Ignores planted instructions, still does the allowed work

4826.6s

4.7K9.1K

81m 40s

Diagnosis & investigation

Chooses diagnostic checks, repairs supported causes and verifies recovery

020.2s

3.2K6.8K

01m 15s

What you'd run

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

SpecADeepSeek V4.1 Flash 1017BMellum2.1 Thinking

Speed

243 tok/s

90 tok/s

Machine

Hosted

Apple M5 Pro, 48 GB

Memory at 32K

—

12.9 GB

Context

128K

Cost per run

$0.75

$2.46 est.

Cost per case

$0.0049

$0.016 est.

Time per case

26.0s

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.

DeepSeek V4.1 Flash 1017Mellum2.1 Thinking

Every test case

153 test cases

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

Agents120

Structured Extraction160

Coding90

Grounded Answers130

Instruction Following120

Business Logic140

Security & permissions101

Diagnosis & investigation00

86 won by DeepSeek V4.1 Flash 10171 won by Mellum2.1 Thinking66 tied