Flint vs MiMo-V2.6-Flash
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. Flint spends 2.5× as many thinking tokens per test case.
TaskAFlintBMiMo-V2.6-Flash
Overall
211m 17s
451m 21s
Agents
1447.7s
381m 19s
Structured Extraction
291m 17s
6238.0s
Coding
141m 19s
242m 40s
Grounded Answers
2957.7s
6242.1s
Instruction Following
352m 28s
641m 2s
Business Logic
101m 20s
4356.4s
Security & permissions
3552.6s
662m 6s
Diagnosis & investigation
057.0s
01m 33s
What you'd run
The build behind each score. Orange marks the faster, smaller, cheaper or longer-context side.
SpecAFlintBMiMo-V2.6-Flash
Speed
149 tok/s (better)
66 tok/s
Machine
Hosted
Hosted
Memory at 32K
—
—
Context
256K
1.1M (better)
Cost per run
Not measured
$0.33
Cost per case
Not measured
$0.0021
Time per case
1m 17s (better)
1m 21s
Quantization
BF16
BF16
Harness
openrouter
openrouter
The race
Each model's running score as the test cases go by. The dashed line is a perfect run.
Every test case
Each square is one test case, colored by the model that scored higher.
Agents
Structured Extraction
Coding
Grounded Answers
Instruction Following
Business Logic
Security & permissions
Diagnosis & investigation