#53 KAT-Coder-Pro V2.5
high- Cost
- $0.009
- Time
- 26.4s
- Tokens
- 3,117 tok
AI BENCHY Compare
Kwaipilot: KAT-Coder-Pro V2.5 vs xAI: Grok 4.3
Summary
The average score is effectively tied at 7.7 vs 7.7. KAT-Coder-Pro V2.5 (high) has the lower benchmark cost at $0.246 vs $0.614. KAT-Coder-Pro V2.5 (high) is faster at 12.30s vs 47.51s, with pass rates of 63.5% vs 71.4%.
Recommended modelKAT-Coder-Pro V2.5 (high)It has the best score here (7.7), while costing about 2.5x less than Grok 4.3 (medium).
Last updated at: 2026-07-14
| Metric | KAT-Coder-Pro V2.5 KAT-Coder-Pro V2.5 high | Grok 4.3 Grok 4.3 medium |
|---|---|---|
| Score | 7.7 | 7.7 |
| Rank | #53 | #48 |
| Reliability | 10.0 | 10.0 |
| Consistency | 8.1 | 8.5 |
| Tests Correct | ||
| Attempt pass rate | 63.5% | 71.4% |
| Flaky tests | 5 | 4 |
| Total Runs | 63 | 63 |
| Cost per result | 2.228 | 4.724 |
| Total Cost | $0.246 | $0.614 |
| Input Price | $0.740 / 1M | $1.250 / 1M |
| Output Price | $2.960 / 1M | $2.500 / 1M |
| Total Input Tokens | 42,750 | 44,472 |
| Output Tokens | 2,509 | 1,981 |
| Reasoning Tokens | 69,594 | 221,382 |
| Response Time (avg) | 12.30s | 47.51s |
| Response Time (max) | 95.65s | 216.69s |
| Response Time (total) | 258.34s | 997.68s |
Generation showcase
Hamster playing table tennis
Prompt: Create a detailed SVG illustration of a hamster playing table tennis.
#48 xAI: Grok 4.3
medium- Cost
- $0.009
- Time
- 19.0s
- Tokens
- 3,661 tok
Top Models by Score
Score vs Total Cost
Response Time (avg)
Score vs Response Time (avg)
Total Output Tokens
Score vs Total Output Tokens
Category Breakdown
| Coding | Score | Consistency | Attempt pass rate | Flaky tests | Tests Correct | Response Time (avg) | Input Tokens | Output Tokens | Reasoning Tokens |
|---|---|---|---|---|---|---|---|---|---|
| KAT-Coder-Pro V2.5 | 6.4 | 7.9 | 44.4% | 1 | 22.00s | 7,893 | 422 | 20,461 | |
| Grok 4.3 | 5.9 | 7.7 | 44.4% | 1 | 41.23s | 8,340 | 1,028 | 31,226 |
Quick Compare
Switch Comparison Pair
GPT-5.6 TerralowvsGrok 4.3mediumStep 3.7 FlashlowvsGrok 4.3mediumClaude Opus 4.8lowvsGrok 4.3mediumGPT-5.6 LunahighvsGrok 4.3mediumKAT-Coder-Pro V2.5highvsStep 3.7 FlashlowClaude Opus 4.6mediumvsKAT-Coder-Pro V2.5highKAT-Coder-Pro V2.5highvsGPT-5.6 TerralowKAT-Coder-Pro V2.5highvsMiniMax M3mediumClaude Opus 4.8lowvsKAT-Coder-Pro V2.5highKAT-Coder-Pro V2.5highvsQwen3.5-122B-A10BmediumKAT-Coder-Pro V2.5highvsGrok Build 0.1mediumKAT-Coder-Pro V2.5highvsKimi K2.6mediumFree Available