GPT-6 Luna vs MiMo-V2.6-Flash, head to head

Side A

Side B

Overall

GPT-6 Luna is clearly better. It wins 43 of the 58 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. MiMo-V2.6-Flash spends 2.8× as many thinking tokens per test case.

Thinking tokensAnswer tokens

TaskAGPT-6 LunaBMiMo-V2.6-FlashGPT-6 LunaMiMo-V2.6-Flash

Overall

6149.1s

1.9K4.2K

451m 21s

Agents

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

561m 18s

1.3K2.2K

381m 19s

Structured Extraction

Turns logs, diffs and records into exact JSON

9514.5s

9992.6K

6238.0s

Coding

Writes code and SQL that pass hidden tests

732m 18s

3.6K9.9K

242m 40s

Grounded Answers

Answers from the given documents with sources, never guesses

6715.6s

1.1K2.6K

6242.1s

Instruction Following

Applies every rule, including ones changed mid-conversation

8432.3s

2K4.8K

641m 2s

Business Logic

Computes billing, scheduling and pricing rules exactly

3743.1s

3.2K4.4K

4356.4s

Security & permissions

Ignores planted instructions, still does the allowed work

6224.5s

1K3.8K

662m 6s

Diagnosis & investigation

Chooses diagnostic checks, repairs supported causes and verifies recovery

1737.7s

1.1K1.5K

01m 33s

What you'd run

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

SpecAGPT-6 LunaBMiMo-V2.6-Flash

Speed

90 tok/s

66 tok/s

Machine

Hosted

Memory at 32K

—

Context

1.1M

Cost per run

$0.10

$0.33

Cost per case

$0.00067

$0.0021

Time per case

49.1s

1m 21s

Quantization

—

BF16

Harness

openrouter

The race

1 lead change

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

GPT-6 LunaMiMo-V2.6-Flash

Every test case

153 test cases

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

Agents73

Structured Extraction81

Coding121

Grounded Answers32

Instruction Following50

Business Logic46

Security & permissions32

Diagnosis & investigation10

43 won by GPT-6 Luna15 won by MiMo-V2.6-Flash95 tied