AI BENCHY
Advertise here
#150

KAT-Coder-Air V2.5

Kwaipilot Release: 2026-07-14 Tested on: 2026-07-16 23:32 kwaipilot/kat-coder-air-v2.5::low
~35B total (~3B active)MoEClosedEstimated
(high) (medium) (low) (none)

Summary

KAT-Coder-Air V2.5 scores 5.4 on AI BENCHY and ranks #150. It has 9.9 reliability, a 45.5% pass rate, $0.041 total cost, and 10.09s average response time.

What makes KAT-Coder-Air V2.5 unique: Its total benchmark cost is unusually low for its score range.

Model facts

Researched on 2026-08-12

Estimated
Parameters
~35B total (~3B active)
Architecture
MoE
Availability
Closed
License
-

Best estimate from public evidence; the vendor did not disclose every value. Weights were announced but were not available for the exact checkpoint at research time.

Parameter source Availability source

Score

5.4

Consistency

7.4

Reliability

9.9

Total Cost (Current Price)

$0.041

Total Output Tokens

52,895

Total Input Tokens

61,085

Input Price

$0.150 / 1M

Output Price

$0.600 / 1M

Tests Correct

Wrong Tests: 15

Attempt pass rate: 45.5%

Flaky tests

7

Flaky tests had mixed outcomes across runs (at least one pass and one fail).

Response Time (avg)

10.09s

Response Time (max): 86.23s

Response Time (total): 222.03s

Wrong answer: 7 Extra formatting: 4 API error: 2 Did not follow instructions: 2

Hamster playing table tennis

Prompt: Create a detailed SVG illustration of a hamster playing table tennis.

#150 KAT-Coder-Air V2.5

low
Cost
$0.004
Time
29.9s
Tokens
6,008 tok

Run history

Tested on Score Reliability Tests Correct Total Cost Compare
2026-08-14 03:16 Re-test 5.4 9.9 $0.046 Compare
2026-07-16 23:32 New test added 5.4 9.9 $0.041 Current run
2026-07-14 11:05 Initial run 6.0 9.7 $0.030 Compare

This run used a different benchmark suite. Keep suite changes in mind when reading historical movement.

Run comparison

RunBenchmark coverageScoreConsistencyReliabilityTests CorrectFlaky testsTotal Output TokensTotal Input TokensTotal CostResponse Time (avg)
2026-07-16 23:32 · New test added66/66 attempts5.47.49.97/22752,89561,085$0.04110.09s
2026-07-14 11:05 · Initial run63/63 attempts6.07.69.77/21639,09142,464$0.0306.47s
Difference—-0.6-0.2+0.20+1+13804+18621+$0.012+3626ms

Benchmark coverage differs: 66/66 attempts (Target: 3 repeats per test) versus 63/63 attempts (Target: 3 repeats per test). Totals and repeat-sensitive metrics are not directly comparable.

These two runs used different benchmark suites, so the deltas reflect both model changes and suite changes.

Price History

Historical pricing data for this model from OpenRouter.

Date Input Price Output Price
2026-07-14 11:25 $0.150 / 1M $0.600 / 1M

Charts

Choose the first model, then click a second model to open a side-by-side page.

Total Output Tokens

Score vs Total Output Tokens

Quick Compare

KAT-Coder-Air V2.5lowvsLaguna S 2.1mediumFree AvailableKAT-Coder-Air V2.5lowvsGPT-5.6 LunanoneKAT-Coder-Air V2.5lowvsDeepSeek V4 Flash 0423noneKAT-Coder-Air V2.5lowvsMiMo-V2.5-PrononeKAT-Coder-Air V2.5lowvsQwen3.8 27BnoneKAT-Coder-Air V2.5lowvsKimi K2.5noneKAT-Coder-Air V2.5lowvsQwen3.6 27BnoneKAT-Coder-Air V2.5lowvsDeepSeek V4 Flash 0731noneKAT-Coder-Air V2.5lowvsNemotron 3.5 LightninghighFree AvailableKAT-Coder-Air V2.5lowvsLaguna S 2.1highFree Available

Category Breakdown

Category Score Consistency Tests Correct
Anti-AI Tricks 7.3 5.8
Coding 3.5 7.3
Combined 6.4 5.8
Data parsing and extraction 6.5 10.0
Domain specific 2.9 7.2
General Intelligence 5.0 10.0
Instructions following 9.8 10.0
Puzzle Solving 3.1 4.6
Tool Calling 10.0 10.0
Trivia 3.0 10.0

Compared models