- Rank
- #87
- Total Output Tokens
- 228,663
- Response Time (avg)
- 91.29s
- Total Cost
- $0.556
DeepSeek V4 Pro 0423 (high) vs Gemini 3.8 Flash (medium)
DeepSeek V4 Pro 0423 (high) leads on average score with 8.1 vs 8.0. DeepSeek V4 Pro 0423 (high) has the lower benchmark cost at $0.556 vs $0.769. Gemini 3.8 Flash (medium) is faster at 16.64s vs 91.29s, with pass rates of 65.2% vs 89.9%.
Compared models
- Rank
- #94
- Total Output Tokens
- 167,486
- Response Time (avg)
- 16.64s
- Total Cost
- $0.769
Recommended model
Gemini 3.8 Flash (medium)
Its score stays close to the best score here (8.0 vs 8.1), while responding about 5.5x faster than DeepSeek V4 Pro 0423 (high).
Detailed comparison
| Metric | DeepSeek V4 Pro 0423 DeepSeek V4 Pro 0423 high | Gemini 3.8 Flash Gemini 3.8 Flash medium |
|---|---|---|
| Score | 8.1 | 8.0 |
| Rank | #87 | #94 |
| Reliability | 10.0 | 9.9 |
| Consistency | 7.8 | 9.0 |
| Attempts | 69/69 | 69/69 |
| Tests Correct | ||
| Attempt pass rate | 65.2% | 89.9% |
| Flaky tests | 6 | 3 |
| Total Runs | 69 | 69 |
| Cost per result | 3.717 | 4.046 |
| Total Cost | $0.556 | $0.769 |
| Input Price | $0.783 / 1M | $0.750 / 1M |
| Output Price | $1.566 / 1M | $3.750 / 1M |
| Total Input Tokens | 287,061 | 187,331 |
| Output Tokens | 32,056 | 7,555 |
| Reasoning Tokens | 196,607 | 159,931 |
| Response Time (avg) | 91.29s | 16.64s |
| Response Time (max) | 416.76s | 108.60s |
| Response Time (total) | 2099.56s | 382.64s |
| Parameters | 1.6T total (49B active) | ~500B total (~20B active) |
| Availability | Open source | Closed |
Model generation showcase
Hamster playing table tennis
Prompt: Create a detailed SVG illustration of a hamster playing table tennis.
#87 DeepSeek V4 Pro 0423
high- Cost
- $0.023
- Time
- 257.6s
- Tokens
- 14,870 tok
#94 Gemini 3.8 Flash
medium- Cost
- $0.037
- Time
- 54.6s
- Tokens
- 9,712 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
| Agentic | Score | Consistency | Attempt pass rate | Flaky tests | Tests Correct | Response Time (avg) | Input Tokens | Output Tokens | Reasoning Tokens |
|---|---|---|---|---|---|---|---|---|---|
| DeepSeek V4 Pro 0423 | 10.0 | 10.0 | 100.0% | 0 | 64.54s | 196,304 | 9,128 | 10,091 | |
| Gemini 3.8 Flash | 2.8 | 1.6 | 33.3% | 1 | 63.72s | 91,538 | 489 | 12,089 |
| Anti-AI Tricks | Score | Consistency | Attempt pass rate | Flaky tests | Tests Correct | Response Time (avg) | Input Tokens | Output Tokens | Reasoning Tokens |
|---|---|---|---|---|---|---|---|---|---|
| DeepSeek V4 Pro 0423 | 5.7 | 5.9 | 58.3% | 2 | 25.70s | 536 | 149 | 3,214 | |
| Gemini 3.8 Flash | 10.0 | 10.0 | 100.0% | 0 | 2.66s | 492 | 145 | 3,131 |
| Coding | Score | Consistency | Attempt pass rate | Flaky tests | Tests Correct | Response Time (avg) | Input Tokens | Output Tokens | Reasoning Tokens |
|---|---|---|---|---|---|---|---|---|---|
| DeepSeek V4 Pro 0423 | 6.3 | 8.7 | 33.3% | 0 | 243.00s | 5,090 | 383 | 84,580 | |
| Gemini 3.8 Flash | 8.0 | 9.9 | 66.7% | 0 | 11.69s | 8,118 | 454 | 19,433 |
| Combined | Score | Consistency | Attempt pass rate | Flaky tests | Tests Correct | Response Time (avg) | Input Tokens | Output Tokens | Reasoning Tokens |
|---|---|---|---|---|---|---|---|---|---|
| DeepSeek V4 Pro 0423 | 10.0 | 10.0 | 100.0% | 0 | 78.99s | 66,082 | 4,582 | 25,404 | |
| Gemini 3.8 Flash | 8.7 | 6.9 | 83.3% | 1 | 60.28s | 71,646 | 5,615 | 59,676 |
| Data parsing and extraction | Score | Consistency | Attempt pass rate | Flaky tests | Tests Correct | Response Time (avg) | Input Tokens | Output Tokens | Reasoning Tokens |
|---|---|---|---|---|---|---|---|---|---|
| DeepSeek V4 Pro 0423 | 10.0 | 10.0 | 100.0% | 0 | 25.03s | 7,690 | 274 | 2,166 | |
| Gemini 3.8 Flash | 10.0 | 10.0 | 100.0% | 0 | 5.74s | 7,548 | 279 | 4,386 |
| Domain specific | Score | Consistency | Attempt pass rate | Flaky tests | Tests Correct | Response Time (avg) | Input Tokens | Output Tokens | Reasoning Tokens |
|---|---|---|---|---|---|---|---|---|---|
| DeepSeek V4 Pro 0423 | 3.6 | 7.2 | 22.2% | 1 | 249.47s | 578 | 16,870 | 58,188 | |
| Gemini 3.8 Flash | 8.2 | 7.2 | 88.9% | 1 | 31.77s | 644 | 12 | 44,973 |
| General Intelligence | Score | Consistency | Attempt pass rate | Flaky tests | Tests Correct | Response Time (avg) | Input Tokens | Output Tokens | Reasoning Tokens |
|---|---|---|---|---|---|---|---|---|---|
| DeepSeek V4 Pro 0423 | 10.0 | 10.0 | 100.0% | 0 | 8.83s | 471 | 115 | 1,013 | |
| Gemini 3.8 Flash | 10.0 | 10.0 | 100.0% | 0 | 3.65s | 486 | 78 | 1,185 |
| Instructions following | Score | Consistency | Attempt pass rate | Flaky tests | Tests Correct | Response Time (avg) | Input Tokens | Output Tokens | Reasoning Tokens |
|---|---|---|---|---|---|---|---|---|---|
| DeepSeek V4 Pro 0423 | 7.8 | 6.6 | 83.3% | 1 | 8.73s | 627 | 66 | 2,726 | |
| Gemini 3.8 Flash | 10.0 | 10.0 | 100.0% | 0 | 4.81s | 615 | 72 | 2,944 |
| Puzzle Solving | Score | Consistency | Attempt pass rate | Flaky tests | Tests Correct | Response Time (avg) | Input Tokens | Output Tokens | Reasoning Tokens |
|---|---|---|---|---|---|---|---|---|---|
| DeepSeek V4 Pro 0423 | 6.9 | 4.9 | 77.8% | 2 | 56.85s | 591 | 178 | 2,563 | |
| Gemini 3.8 Flash | 10.0 | 10.0 | 100.0% | 0 | 2.73s | 558 | 151 | 2,720 |
| Tool Calling | Score | Consistency | Attempt pass rate | Flaky tests | Tests Correct | Response Time (avg) | Input Tokens | Output Tokens | Reasoning Tokens |
|---|---|---|---|---|---|---|---|---|---|
| DeepSeek V4 Pro 0423 | 9.8 | 10.0 | 100.0% | 0 | 15.92s | 8,909 | 295 | 701 | |
| Gemini 3.8 Flash | 10.0 | 10.0 | 100.0% | 0 | 18.47s | 5,530 | 249 | 6,512 |
| Trivia | Score | Consistency | Attempt pass rate | Flaky tests | Tests Correct | Response Time (avg) | Input Tokens | Output Tokens | Reasoning Tokens |
|---|---|---|---|---|---|---|---|---|---|
| DeepSeek V4 Pro 0423 | 3.0 | 10.0 | 0.0% | 0 | 34.01s | 183 | 16 | 5,961 | |
| Gemini 3.8 Flash | 10.0 | 10.0 | 100.0% | 0 | 5.94s | 156 | 11 | 2,882 |
Quick Compare
Switch Comparison Pair
Claude Opus 4.7mediumvsDeepSeek V4 Pro 0423highGemini 3.8 FlashmediumvsGPT-5.6 TerralowGemini 3.8 FlashmediumvsGPT-5.6 LunahighGemini 3.8 FlashmediumvsGLM 5.3 FlashXmaxClaude Fable 5.1highvsGemini 3.8 FlashmediumSeed-2.0-CodelowvsDeepSeek V4 Pro 0423highGemini 3.8 FlashmediumvsQwen3.7 MaxnoneDeepSeek V4.1 FlashlowvsGemini 3.8 FlashmediumSeed-2.0-LitemediumvsDeepSeek V4 Pro 0423highDeepSeek V4 Pro 0423highvsNemotron 3 UltramediumFree AvailableDeepSeek V4 Flash 0423highvsGemini 3.8 FlashmediumGemini 3.8 FlashmediumvsQwen3.8 2.4T A95Bhigh