#257 KAT-Coder-Air V2.5
none- Cost
- $0.002
- Time
- 14.5s
- Tokens
- 2,619 tok
Kwaipilot
Release: 2026-07-14
Tested on: 2026-08-14 03:15
kwaipilot/kat-coder-air-v2.5::none
~35B total (~3B active)MoEClosedEstimated
Summary
KAT-Coder-Air V2.5 scores 4.8 on AI BENCHY and ranks #257. It has 10.0 reliability, a 36.4% pass rate, $0.070 total cost, and 12.46s average response time.
What makes KAT-Coder-Air V2.5 unique: It stands out most in Instructions following, where it ranks #1, while Puzzle Solving is its weakest area at #12.
Model facts
Researched on 2026-08-12
- 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.
4.8
Consistency
7.2
10.0
$0.070
Total Output Tokens
97,752
Total Input Tokens
69,376
Input Price
$0.150 / 1M
Output Price
$0.600 / 1M
Flaky tests
7
Flaky tests had mixed outcomes across runs (at least one pass and one fail).
Hamster playing table tennis
Prompt: Create a detailed SVG illustration of a hamster playing table tennis.
Run history
| Tested on | Score | Reliability | Tests Correct | Total Cost | Compare |
|---|---|---|---|---|---|
| 2026-08-14 03:15 Re-test | 4.8 | 10.0 | $0.070 | Current run | |
| 2026-07-16 23:28 New test added | 4.8 | 10.0 | $0.067 | Compare | |
| 2026-07-14 11:05 Initial run | 4.9 | 10.0 | $0.037 | Compare |
This run used a different benchmark suite. Keep suite changes in mind when reading historical movement.
Run comparison
| Run | Benchmark coverage | Score | Consistency | Reliability | Tests Correct | Flaky tests | Total Output Tokens | Total Input Tokens | Total Cost | Response Time (avg) |
|---|---|---|---|---|---|---|---|---|---|---|
| 2026-08-14 03:15 · Current run | 66/66 attempts | 4.8 | 7.2 | 10.0 | 5/22 | 7 | 97,752 | 69,376 | $0.070 | 12.46s |
| 2026-07-14 11:05 · Initial run | 63/63 attempts | 4.9 | 7.1 | 10.0 | 5/21 | 7 | 49,933 | 41,328 | $0.037 | 6.99s |
| Difference | — | -0.1 | +0.1 | 0.0 | 0 | 0 | +47819 | +28048 | +$0.033 | +5470ms |
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 |
|---|---|---|
| 2026-08-14 12:51 | $0.150 / 1M | $0.600 / 1M |
Charts
Choose the first model, then click a second model to open a side-by-side page.
Top Models by Score
Score vs Total Cost
Response Time (avg)
Score vs Response Time (avg)
Total Output Tokens
Score vs Total Output Tokens
Quick Compare
KAT-Coder-Air V2.5nonevsTrinity Large ThinkinghighKAT-Coder-Air V2.5nonevsGLM 4.7 FlashnoneKAT-Coder-Air V2.5nonevsGPT-5.4 NanononeKAT-Coder-Air V2.5nonevsTrinity Large PreviewnoneKAT-Coder-Air V2.5nonevsNemotron 3.5 LightninglowFree AvailableKAT-Coder-Air V2.5nonevsCobuddymediumKAT-Coder-Air V2.5nonevsRing-2.6-1TnoneKAT-Coder-Air V2.5nonevsMercury 2noneKAT-Coder-Air V2.5nonevsNemotron 3 SupernoneFree AvailableKAT-Coder-Air V2.5nonevsQwen3 Coder Nextmedium
Category Breakdown
| Category | Score | Consistency | Tests Correct |
|---|---|---|---|
| Anti-AI Tricks | 5.3 | 5.6 | |
| Coding | 3.3 | 9.6 | |
| Combined | 3.8 | 1.6 | |
| Data parsing and extraction | 6.3 | 5.8 | |
| Domain specific | 3.0 | 10.0 | |
| General Intelligence | 5.0 | 10.0 | |
| Instructions following | 9.9 | 10.0 | |
| Puzzle Solving | 2.9 | 4.4 | |
| Tool Calling | 10.0 | 10.0 | |
| Trivia | 3.0 | 10.0 |