Flint vs Qwen3.8 Flash Next, head to head

Side A

Side B

Overall

Qwen3.8 Flash Next is clearly better. It wins 50 of the 65 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. Flint spends 3.0× as many thinking tokens per test case.

Thinking tokensAnswer tokens

TaskAFlintBQwen3.8 Flash NextFlintQwen3.8 Flash Next

Overall

211m 17s

10.9K4.1K

421m 15s

Agents

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

1447.7s

4.8K4.6K

522m 4s

Structured Extraction

Turns logs, diffs and records into exact JSON

291m 17s

10.9K3.1K

5042.2s

Coding

Writes code and SQL that pass hidden tests

141m 19s

12K6.2K

521m 36s

Grounded Answers

Answers from the given documents with sources, never guesses

2957.7s

8.7K3.1K

5147.0s

Instruction Following

Applies every rule, including ones changed mid-conversation

352m 28s

22.2K5.2K

621m 30s

Business Logic

Computes billing, scheduling and pricing rules exactly

101m 20s

11.7K3.9K

251m 0s

Security & permissions

Ignores planted instructions, still does the allowed work

3552.6s

7.1K2.8K

451m 8s

Diagnosis & investigation

Chooses diagnostic checks, repairs supported causes and verifies recovery

057.0s

7.3K3.4K

01m 15s

What you'd run

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

SpecAFlintBQwen3.8 Flash Next

Speed

149 tok/s

68 tok/s

Machine

Hosted

Memory at 32K

—

Context

256K

Cost per run

Not measured

$0.46

Cost per case

Not measured

$0.003

Time per case

1m 17s

1m 15s

Quantization

BF16

—

Harness

openrouter

The race

No lead changes

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

FlintQwen3.8 Flash Next

Every test case

153 test cases

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

Agents09

Structured Extraction68

Coding010

Grounded Answers16

Instruction Following47

Business Logic15

Security & permissions35

Diagnosis & investigation00

15 won by Flint50 won by Qwen3.8 Flash Next88 tied