- Rank
- #126
- Total Output Tokens
- 550,855
- Response Time (avg)
- 84.66s
- Total Cost
- $0.066
DeepSeek V4 Flash 0731 (medium) vs KAT-Coder-Pro V2.5 (high)
DeepSeek V4 Flash 0731 (medium) leads on average score with 7.3 vs 7.2. DeepSeek V4 Flash 0731 (medium) has the lower benchmark cost at $0.066 vs $0.506. KAT-Coder-Pro V2.5 (high) is faster at 22.11s vs 84.66s, with pass rates of 71.2% vs 63.6%.
Compared models
- Rank
- #129
- Total Output Tokens
- 144,359
- Response Time (avg)
- 22.11s
- Total Cost
- $0.506
Recommended model
DeepSeek V4 Flash 0731 (medium)
It has the best score here (7.3), while costing about 7.7x less than KAT-Coder-Pro V2.5 (high).
Detailed comparison
| Metric | DeepSeek V4 Flash 0731 DeepSeek V4 Flash 0731 medium | KAT-Coder-Pro V2.5 KAT-Coder-Pro V2.5 high |
|---|---|---|
| Score | 7.3 | 7.2 |
| Rank | #126 | #129 |
| Reliability | 10.0 | 10.0 |
| Consistency | 7.1 | 7.8 |
| Attempts | 66/66 | 66/66 |
| Tests Correct | ||
| Attempt pass rate | 71.2% | 63.6% |
| Flaky tests | 8 | 6 |
| Total Runs | 66 | 66 |
| Cost per result | 0.960 | 4.599 |
| Total Cost | $0.066 | $0.506 |
| Input Price | $0.065 / 1M | $0.740 / 1M |
| Output Price | $0.180 / 1M | $2.960 / 1M |
| Total Input Tokens | 89,819 | 106,085 |
| Output Tokens | 58,398 | 9,071 |
| Reasoning Tokens | 492,457 | 135,288 |
| Response Time (avg) | 84.66s | 22.11s |
| Response Time (max) | 288.78s | 199.97s |
| Response Time (total) | 1862.46s | 486.31s |
| Parameters | 284B total (13B active) | ~700B total (~72B active) |
| Availability | Closed | Closed |
Model generation showcase
Hamster playing table tennis
Prompt: Create a detailed SVG illustration of a hamster playing table tennis.
#126 DeepSeek V4 Flash 0731
medium- Cost
- $0.004
- Time
- 93.3s
- Tokens
- 13,252 tok
#129 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
| Anti-AI Tricks | Score | Consistency | Attempt pass rate | Flaky tests | Tests Correct | Response Time (avg) | Input Tokens | Output Tokens | Reasoning Tokens |
|---|---|---|---|---|---|---|---|---|---|
| DeepSeek V4 Flash 0731 | 8.2 | 7.9 | 83.3% | 1 | 17.95s | 698 | 2,101 | 6,715 | |
| KAT-Coder-Pro V2.5 | 7.0 | 8.0 | 66.7% | 1 | 3.17s | 672 | 447 | 2,716 |
| Coding | Score | Consistency | Attempt pass rate | Flaky tests | Tests Correct | Response Time (avg) | Input Tokens | Output Tokens | Reasoning Tokens |
|---|---|---|---|---|---|---|---|---|---|
| DeepSeek V4 Flash 0731 | 6.4 | 4.4 | 77.8% | 2 | 198.78s | 6,032 | 4,109 | 160,154 | |
| KAT-Coder-Pro V2.5 | 6.4 | 7.9 | 44.4% | 1 | 22.00s | 7,893 | 422 | 20,461 |
| Combined | Score | Consistency | Attempt pass rate | Flaky tests | Tests Correct | Response Time (avg) | Input Tokens | Output Tokens | Reasoning Tokens |
|---|---|---|---|---|---|---|---|---|---|
| DeepSeek V4 Flash 0731 | 7.3 | 5.8 | 83.3% | 1 | 150.58s | 65,118 | 40,110 | 97,296 | |
| KAT-Coder-Pro V2.5 | 7.3 | 5.8 | 83.3% | 1 | 106.70s | 78,458 | 6,956 | 60,306 |
| Data parsing and extraction | Score | Consistency | Attempt pass rate | Flaky tests | Tests Correct | Response Time (avg) | Input Tokens | Output Tokens | Reasoning Tokens |
|---|---|---|---|---|---|---|---|---|---|
| DeepSeek V4 Flash 0731 | 10.0 | 10.0 | 100.0% | 0 | 5.21s | 7,290 | 287 | 1,519 | |
| KAT-Coder-Pro V2.5 | 10.0 | 10.0 | 100.0% | 0 | 4.13s | 7,776 | 278 | 1,960 |
| Domain specific | Score | Consistency | Attempt pass rate | Flaky tests | Tests Correct | Response Time (avg) | Input Tokens | Output Tokens | Reasoning Tokens |
|---|---|---|---|---|---|---|---|---|---|
| DeepSeek V4 Flash 0731 | 2.9 | 7.2 | 11.1% | 1 | 203.11s | 592 | 9,333 | 154,069 | |
| KAT-Coder-Pro V2.5 | 2.9 | 4.4 | 22.2% | 2 | 44.66s | 780 | 33 | 43,896 |
| General Intelligence | Score | Consistency | Attempt pass rate | Flaky tests | Tests Correct | Response Time (avg) | Input Tokens | Output Tokens | Reasoning Tokens |
|---|---|---|---|---|---|---|---|---|---|
| DeepSeek V4 Flash 0731 | 6.1 | 3.1 | 66.7% | 1 | 12.16s | 471 | 66 | 1,589 | |
| KAT-Coder-Pro V2.5 | 5.1 | 10.0 | 0.0% | 0 | 3.27s | 516 | 139 | 647 |
| Instructions following | Score | Consistency | Attempt pass rate | Flaky tests | Tests Correct | Response Time (avg) | Input Tokens | Output Tokens | Reasoning Tokens |
|---|---|---|---|---|---|---|---|---|---|
| DeepSeek V4 Flash 0731 | 8.5 | 6.8 | 83.3% | 1 | 2.42s | 706 | 109 | 580 | |
| KAT-Coder-Pro V2.5 | 9.9 | 10.0 | 100.0% | 0 | 2.67s | 699 | 84 | 976 |
| Puzzle Solving | Score | Consistency | Attempt pass rate | Flaky tests | Tests Correct | Response Time (avg) | Input Tokens | Output Tokens | Reasoning Tokens |
|---|---|---|---|---|---|---|---|---|---|
| DeepSeek V4 Flash 0731 | 8.2 | 7.3 | 88.9% | 1 | 65.12s | 594 | 1,530 | 66,970 | |
| KAT-Coder-Pro V2.5 | 8.2 | 7.2 | 88.9% | 1 | 3.38s | 696 | 428 | 1,936 |
| Tool Calling | Score | Consistency | Attempt pass rate | Flaky tests | Tests Correct | Response Time (avg) | Input Tokens | Output Tokens | Reasoning Tokens |
|---|---|---|---|---|---|---|---|---|---|
| DeepSeek V4 Flash 0731 | 10.0 | 10.0 | 100.0% | 0 | 4.41s | 8,178 | 374 | 152 | |
| KAT-Coder-Pro V2.5 | 10.0 | 10.0 | 100.0% | 0 | 28.04s | 8,391 | 268 | 896 |
| Trivia | Score | Consistency | Attempt pass rate | Flaky tests | Tests Correct | Response Time (avg) | Input Tokens | Output Tokens | Reasoning Tokens |
|---|---|---|---|---|---|---|---|---|---|
| DeepSeek V4 Flash 0731 | 3.0 | 10.0 | 0.0% | 0 | 56.64s | 140 | 379 | 3,413 | |
| KAT-Coder-Pro V2.5 | 3.0 | 10.0 | 0.0% | 0 | 5.22s | 204 | 16 | 1,494 |
Quick Compare
Switch Comparison Pair
Claude Opus 4.8nonevsDeepSeek V4 Flash 0731mediumKAT-Coder-Pro V2.5highvsKimi K2.6mediumDeepSeek V4 Flash 0731mediumvsQwen3.7 PlusnoneDeepSeek V4 Flash 0731mediumvsGLM 5.3 FlashhighKAT-Coder-Pro V2.5highvsQwen3.5 Plus 2026-04-20mediumClaude Opus 4.8nonevsKAT-Coder-Pro V2.5highKAT-Coder-Pro V2.5highvsQwen3.7 FlashlowDeepSeek V4 Flash 0731mediumvsKAT-Coder-Pro V2.5lowKAT-Coder-Pro V2.5highvsMuse Glimmer 30BmediumDeepSeek V4 Flash 0731mediumvsGLM 5.3highKAT-Coder-Pro V2.5highvsQwen3.7 PlusnoneClaude Sonnet 4.6nonevsDeepSeek V4 Flash 0731medium