#55 DeepSeek V4 Pro
high- Cost
- $0.023
- Time
- 257.6s
- Tokens
- 14,870 tok
AI BENCHY Compare
DeepSeek: DeepSeek V4 Pro vs Kwaipilot: KAT-Coder-Pro V2.5
Summary
KAT-Coder-Pro V2.5 (high) leads on average score with 7.7 vs 7.6. DeepSeek V4 Pro (high) has the lower benchmark cost at $0.157 vs $0.246. KAT-Coder-Pro V2.5 (high) is faster at 12.30s vs 77.20s, with pass rates of 66.7% vs 63.5%.
Recommended modelKAT-Coder-Pro V2.5 (high)It has the best score here (7.7), while responding about 6.3x faster than DeepSeek V4 Pro (high).
Last updated at: 2026-07-14
| Metric | DeepSeek V4 Pro DeepSeek V4 Pro high | KAT-Coder-Pro V2.5 KAT-Coder-Pro V2.5 high |
|---|---|---|
| Score | 7.6 | 7.7 |
| Rank | #55 | #53 |
| Reliability | 9.3 | 10.0 |
| Consistency | 7.0 | 8.1 |
| Tests Correct | ||
| Attempt pass rate | 66.7% | 63.5% |
| Flaky tests | 8 | 5 |
| Total Runs | 63 | 63 |
| Cost per result | 1.742 | 2.228 |
| Total Cost | $0.157 | $0.246 |
| Input Price | $0.435 / 1M | $0.740 / 1M |
| Output Price | $0.870 / 1M | $2.960 / 1M |
| Total Input Tokens | 38,726 | 42,750 |
| Output Tokens | 6,334 | 2,509 |
| Reasoning Tokens | 159,151 | 69,594 |
| Response Time (avg) | 77.20s | 12.30s |
| Response Time (max) | 416.76s | 95.65s |
| Response Time (total) | 1621.17s | 258.34s |
Generation showcase
Hamster playing table tennis
Prompt: Create a detailed SVG illustration of a hamster playing table tennis.
#53 KAT-Coder-Pro V2.5
high- Cost
- $0.009
- Time
- 26.4s
- Tokens
- 3,117 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 |
|---|---|---|---|---|---|---|---|---|---|
| DeepSeek V4 Pro | 6.1 | 4.6 | 66.7% | 2 | 243.00s | 5,090 | 383 | 84,580 | |
| KAT-Coder-Pro V2.5 | 6.4 | 7.9 | 44.4% | 1 | 22.00s | 7,893 | 422 | 20,461 |
Quick Compare
Switch Comparison Pair
DeepSeek V4 ProhighvsMiniMax M3mediumDeepSeek V4 ProhighvsGrok Build 0.1mediumKAT-Coder-Pro V2.5highvsStep 3.7 FlashlowClaude Opus 4.6mediumvsKAT-Coder-Pro V2.5highDeepSeek V4 ProhighvsGPT-5.6 LunamediumKAT-Coder-Pro V2.5highvsGPT-5.6 TerralowKAT-Coder-Pro V2.5highvsGrok 4.3mediumKAT-Coder-Pro V2.5highvsMiniMax M3mediumDeepSeek V4 ProhighvsKimi K2.5mediumDeepSeek V4 ProhighvsMercury 2mediumClaude Opus 4.8lowvsKAT-Coder-Pro V2.5highKAT-Coder-Pro V2.5highvsQwen3.5-122B-A10Bmedium