Qwen3.8 27B (high) leads on average score with 8.4 vs 8.3. Qwen3.8 27B (high) has the lower benchmark cost at ~$0.287 vs $0.537. DeepSeek V4.1 Flash (max) is faster at 61.35s vs 167.89s, with pass rates of 68.2% vs 77.3%.
Last updated at:
2026-09-10
Compared models
Rank
#62
Total Output Tokens
871,126
Response Time (avg)
61.35s
Total Cost
$0.537
Candidate responses were generated on local hardware. OpenRouter was used only for judging.
Rank
#59
Total Output Tokens
656,635
Response Time (avg)
167.89s
Total Cost
~$0.287Estimated GPU electricity cost only. It excludes the rest of the computer and hardware ownership. NVIDIA GeForce RTX 3090: 2.8999 h × 0.300 kW × €0.2896/kWh × 1.138 USD/EUR = ~$0.286715.
Recommended modelQwen3.8 27B (high)
It has the best score here (8.4), while costing about 1.9x less than DeepSeek V4.1 Flash (max).
Qwen3.8 27BQwen3.8 27BhighCandidate responses were generated on local hardware. OpenRouter was used only for judging.Release: 2026-08-14
Score
8.3Summarizes broad quality across our full private benchmark suite, so ranking reflects consistent performance.…
8.4Summarizes broad quality across our full private benchmark suite, so ranking reflects consistent performance.…
Rank
#62
#59
Reliability
9.1First-attempt success score: 10.0 means no retryable target API or rate-limit failures before successful calls; tracked failures lower the score.…
9.6First-attempt success score: 10.0 means no retryable target API or rate-limit failures before successful calls; tracked failures lower the score.…
Consistency
9.2Consistency score reflects run-to-run stability (10 = very consistent, even if consistently wrong).…
9.2Consistency score reflects run-to-run stability (10 = very consistent, even if consistently wrong).…
Attempts
66/66
66/66
Tests Correct
A test is fully passed only if every run passed for that test.Wrong answer: 6API error: 1Did not follow instructions: 1Response Time (avg)61.35sResponse Time (max)335.08sResponse Time (total)1349.61sA test is fully passed only if every run passed for that test.…
A test is fully passed only if every run passed for that test.Did not follow instructions: 2No answer: 2Wrong answer: 2Response Time (avg)167.89sResponse Time (max)640.85sResponse Time (total)3693.54sA test is fully passed only if every run passed for that test.…
Attempt pass rate
68.2%Attempt pass rate = passed attempts / total attempts across runs.…
77.3%Attempt pass rate = passed attempts / total attempts across runs.…
Flaky tests
2Flaky tests had mixed outcomes across runs (at least one pass and one fail).…
2Flaky tests had mixed outcomes across runs (at least one pass and one fail).…
Total Runs
66Total Runs…
66Total Runs…
Cost per result
3.829
~1.792Estimated GPU electricity cost only. It excludes the rest of the computer and hardware ownership. NVIDIA GeForce RTX 3090: 2.8999 h × 0.300 kW × €0.2896/kWh × 1.138 USD/EUR = ~$0.286715.
Total Cost
$0.537
~$0.287Estimated GPU electricity cost only. It excludes the rest of the computer and hardware ownership. NVIDIA GeForce RTX 3090: 2.8999 h × 0.300 kW × €0.2896/kWh × 1.138 USD/EUR = ~$0.286715.
Input Price
$0.150 / 1MInput Price…
N/AInput Price…
Output Price
$0.600 / 1MOutput Price…
N/AOutput Price…
Total Input Tokens
89,024Total Input Tokens…
100,179Total Input Tokens…
Output Tokens
5,558Output Tokens…
904Output Tokens…
Reasoning Tokens
865,568Reasoning Tokens…
655,731Reasoning Tokens…
Response Time (avg)
61.35sResponse Time (avg)…
167.89sResponse Time (avg)…
Response Time (max)
335.08sResponse Time (max)…
640.85sResponse Time (max)…
Response Time (total)
1349.61sResponse Time (total)…
3693.54sResponse Time (total)…
Parameters
748B total (16B active)
27.3B
Availability
Open source
Weights available
Model generation showcase
Hamster playing table tennis
Prompt: Create a detailed SVG illustration of a hamster playing table tennis.
#62 DeepSeek V4.1 Flash
max
Cost
$0.066
Time
211.0s
Tokens
54,712 tok
#59 Qwen3.8 27B
highCandidate responses were generated on local hardware. OpenRouter was used only for judging.
Cost
~$0.008Estimated GPU electricity cost only. It excludes the rest of the computer and hardware ownership. NVIDIA GeForce RTX 3090: 0.0750 h × 0.300 kW × €0.2896/kWh × 1.138 USD/EUR = ~$0.007419.
8.3Summarizes broad quality across our full private benchmark suite, so ranking reflects consistent performance.…
10.0Consistency score reflects run-to-run stability (10 = very consistent, even if consistently wrong).…
75.0%Attempt pass rate = passed attempts / total attempts across runs.…
0Flaky tests had mixed outcomes across runs (at least one pass and one fail).…
A test is fully passed only if every run passed for that test.Wrong answer: 1Response Time (avg)4.35sResponse Time (max)11.86sResponse Time (total)17.41sA test is fully passed only if every run passed for that test.…
4.35sResponse Time (avg)…
852Total Input Tokens…
213Output Tokens…
7,262Reasoning Tokens…
Qwen3.8 27BCandidate responses were generated on local hardware. OpenRouter was used only for judging.
10.0Summarizes broad quality across our full private benchmark suite, so ranking reflects consistent performance.…
10.0Consistency score reflects run-to-run stability (10 = very consistent, even if consistently wrong).…
100.0%Attempt pass rate = passed attempts / total attempts across runs.…
0Flaky tests had mixed outcomes across runs (at least one pass and one fail).…
A test is fully passed only if every run passed for that test.No failed answers.Response Time (avg)8.10sResponse Time (max)21.48sResponse Time (total)32.40sA test is fully passed only if every run passed for that test.…
10.0Summarizes broad quality across our full private benchmark suite, so ranking reflects consistent performance.…
10.0Consistency score reflects run-to-run stability (10 = very consistent, even if consistently wrong).…
100.0%Attempt pass rate = passed attempts / total attempts across runs.…
0Flaky tests had mixed outcomes across runs (at least one pass and one fail).…
A test is fully passed only if every run passed for that test.No failed answers.Response Time (avg)62.50sResponse Time (max)80.58sResponse Time (total)187.50sA test is fully passed only if every run passed for that test.…
62.50sResponse Time (avg)…
7,509Total Input Tokens…
373Output Tokens…
132,357Reasoning Tokens…
Qwen3.8 27BCandidate responses were generated on local hardware. OpenRouter was used only for judging.
8.1Summarizes broad quality across our full private benchmark suite, so ranking reflects consistent performance.…
7.0Consistency score reflects run-to-run stability (10 = very consistent, even if consistently wrong).…
88.9%Attempt pass rate = passed attempts / total attempts across runs.…
1Flaky tests had mixed outcomes across runs (at least one pass and one fail).…
A test is fully passed only if every run passed for that test.Did not follow instructions: 1Response Time (avg)248.62sResponse Time (max)458.25sResponse Time (total)745.85sA test is fully passed only if every run passed for that test.…
10.0Summarizes broad quality across our full private benchmark suite, so ranking reflects consistent performance.…
10.0Consistency score reflects run-to-run stability (10 = very consistent, even if consistently wrong).…
100.0%Attempt pass rate = passed attempts / total attempts across runs.…
0Flaky tests had mixed outcomes across runs (at least one pass and one fail).…
A test is fully passed only if every run passed for that test.No failed answers.Response Time (avg)160.31sResponse Time (max)164.09sResponse Time (total)320.63sA test is fully passed only if every run passed for that test.…
160.31sResponse Time (avg)…
66,677Total Input Tokens…
4,031Output Tokens…
221,091Reasoning Tokens…
Qwen3.8 27BCandidate responses were generated on local hardware. OpenRouter was used only for judging.
10.0Summarizes broad quality across our full private benchmark suite, so ranking reflects consistent performance.…
10.0Consistency score reflects run-to-run stability (10 = very consistent, even if consistently wrong).…
100.0%Attempt pass rate = passed attempts / total attempts across runs.…
0Flaky tests had mixed outcomes across runs (at least one pass and one fail).…
A test is fully passed only if every run passed for that test.No failed answers.Response Time (avg)185.57sResponse Time (max)333.21sResponse Time (total)371.15sA test is fully passed only if every run passed for that test.…
6.5Summarizes broad quality across our full private benchmark suite, so ranking reflects consistent performance.…
10.0Consistency score reflects run-to-run stability (10 = very consistent, even if consistently wrong).…
50.0%Attempt pass rate = passed attempts / total attempts across runs.…
0Flaky tests had mixed outcomes across runs (at least one pass and one fail).…
A test is fully passed only if every run passed for that test.API error: 1Response Time (avg)4.56sResponse Time (max)8.91sResponse Time (total)9.12sA test is fully passed only if every run passed for that test.…
4.56sResponse Time (avg)…
2,397Total Input Tokens…
120Output Tokens…
2,047Reasoning Tokens…
Qwen3.8 27BCandidate responses were generated on local hardware. OpenRouter was used only for judging.
10.0Summarizes broad quality across our full private benchmark suite, so ranking reflects consistent performance.…
10.0Consistency score reflects run-to-run stability (10 = very consistent, even if consistently wrong).…
100.0%Attempt pass rate = passed attempts / total attempts across runs.…
0Flaky tests had mixed outcomes across runs (at least one pass and one fail).…
A test is fully passed only if every run passed for that test.No failed answers.Response Time (avg)34.85sResponse Time (max)57.71sResponse Time (total)69.71sA test is fully passed only if every run passed for that test.…
3.6Summarizes broad quality across our full private benchmark suite, so ranking reflects consistent performance.…
7.2Consistency score reflects run-to-run stability (10 = very consistent, even if consistently wrong).…
22.2%Attempt pass rate = passed attempts / total attempts across runs.…
1Flaky tests had mixed outcomes across runs (at least one pass and one fail).…
A test is fully passed only if every run passed for that test.Wrong answer: 3Response Time (avg)198.41sResponse Time (max)335.08sResponse Time (total)595.23sA test is fully passed only if every run passed for that test.…
198.41sResponse Time (avg)…
909Total Input Tokens…
27Output Tokens…
384,462Reasoning Tokens…
Qwen3.8 27BCandidate responses were generated on local hardware. OpenRouter was used only for judging.
5.3Summarizes broad quality across our full private benchmark suite, so ranking reflects consistent performance.…
10.0Consistency score reflects run-to-run stability (10 = very consistent, even if consistently wrong).…
33.3%Attempt pass rate = passed attempts / total attempts across runs.…
0Flaky tests had mixed outcomes across runs (at least one pass and one fail).…
A test is fully passed only if every run passed for that test.No answer: 1Wrong answer: 1Response Time (avg)526.74sResponse Time (max)640.85sResponse Time (total)1580.21sA test is fully passed only if every run passed for that test.…
10.0Summarizes broad quality across our full private benchmark suite, so ranking reflects consistent performance.…
10.0Consistency score reflects run-to-run stability (10 = very consistent, even if consistently wrong).…
100.0%Attempt pass rate = passed attempts / total attempts across runs.…
0Flaky tests had mixed outcomes across runs (at least one pass and one fail).…
A test is fully passed only if every run passed for that test.No failed answers.Response Time (avg)4.00sResponse Time (max)4.00sResponse Time (total)4.00sA test is fully passed only if every run passed for that test.…
4.00sResponse Time (avg)…
549Total Input Tokens…
148Output Tokens…
1,093Reasoning Tokens…
Qwen3.8 27BCandidate responses were generated on local hardware. OpenRouter was used only for judging.
5.0Summarizes broad quality across our full private benchmark suite, so ranking reflects consistent performance.…
10.0Consistency score reflects run-to-run stability (10 = very consistent, even if consistently wrong).…
0.0%Attempt pass rate = passed attempts / total attempts across runs.…
0Flaky tests had mixed outcomes across runs (at least one pass and one fail).…
A test is fully passed only if every run passed for that test.Did not follow instructions: 1Response Time (avg)8.45sResponse Time (max)8.45sResponse Time (total)8.45sA test is fully passed only if every run passed for that test.…
10.0Summarizes broad quality across our full private benchmark suite, so ranking reflects consistent performance.…
10.0Consistency score reflects run-to-run stability (10 = very consistent, even if consistently wrong).…
100.0%Attempt pass rate = passed attempts / total attempts across runs.…
0Flaky tests had mixed outcomes across runs (at least one pass and one fail).…
A test is fully passed only if every run passed for that test.No failed answers.Response Time (avg)7.64sResponse Time (max)11.30sResponse Time (total)15.29sA test is fully passed only if every run passed for that test.…
7.64sResponse Time (avg)…
783Total Input Tokens…
63Output Tokens…
2,293Reasoning Tokens…
Qwen3.8 27BCandidate responses were generated on local hardware. OpenRouter was used only for judging.
9.8Summarizes broad quality across our full private benchmark suite, so ranking reflects consistent performance.…
10.0Consistency score reflects run-to-run stability (10 = very consistent, even if consistently wrong).…
100.0%Attempt pass rate = passed attempts / total attempts across runs.…
0Flaky tests had mixed outcomes across runs (at least one pass and one fail).…
A test is fully passed only if every run passed for that test.No failed answers.Response Time (avg)8.24sResponse Time (max)10.18sResponse Time (total)16.49sA test is fully passed only if every run passed for that test.…
5.5Summarizes broad quality across our full private benchmark suite, so ranking reflects consistent performance.…
7.1Consistency score reflects run-to-run stability (10 = very consistent, even if consistently wrong).…
44.4%Attempt pass rate = passed attempts / total attempts across runs.…
1Flaky tests had mixed outcomes across runs (at least one pass and one fail).…
A test is fully passed only if every run passed for that test.Did not follow instructions: 1Wrong answer: 1Response Time (avg)30.82sResponse Time (max)73.06sResponse Time (total)92.46sA test is fully passed only if every run passed for that test.…
30.82sResponse Time (avg)…
828Total Input Tokens…
339Output Tokens…
54,014Reasoning Tokens…
Qwen3.8 27BCandidate responses were generated on local hardware. OpenRouter was used only for judging.
7.6Summarizes broad quality across our full private benchmark suite, so ranking reflects consistent performance.…
7.2Consistency score reflects run-to-run stability (10 = very consistent, even if consistently wrong).…
77.8%Attempt pass rate = passed attempts / total attempts across runs.…
1Flaky tests had mixed outcomes across runs (at least one pass and one fail).…
A test is fully passed only if every run passed for that test.No answer: 1Response Time (avg)199.90sResponse Time (max)569.33sResponse Time (total)599.69sA test is fully passed only if every run passed for that test.…
10.0Summarizes broad quality across our full private benchmark suite, so ranking reflects consistent performance.…
10.0Consistency score reflects run-to-run stability (10 = very consistent, even if consistently wrong).…
100.0%Attempt pass rate = passed attempts / total attempts across runs.…
0Flaky tests had mixed outcomes across runs (at least one pass and one fail).…
A test is fully passed only if every run passed for that test.No failed answers.Response Time (avg)9.49sResponse Time (max)9.49sResponse Time (total)9.49sA test is fully passed only if every run passed for that test.…
9.49sResponse Time (avg)…
8,259Total Input Tokens…
228Output Tokens…
693Reasoning Tokens…
Qwen3.8 27BCandidate responses were generated on local hardware. OpenRouter was used only for judging.
10.0Summarizes broad quality across our full private benchmark suite, so ranking reflects consistent performance.…
10.0Consistency score reflects run-to-run stability (10 = very consistent, even if consistently wrong).…
100.0%Attempt pass rate = passed attempts / total attempts across runs.…
0Flaky tests had mixed outcomes across runs (at least one pass and one fail).…
A test is fully passed only if every run passed for that test.No failed answers.Response Time (avg)5.62sResponse Time (max)5.62sResponse Time (total)5.62sA test is fully passed only if every run passed for that test.…
3.0Summarizes broad quality across our full private benchmark suite, so ranking reflects consistent performance.…
10.0Consistency score reflects run-to-run stability (10 = very consistent, even if consistently wrong).…
0.0%Attempt pass rate = passed attempts / total attempts across runs.…
0Flaky tests had mixed outcomes across runs (at least one pass and one fail).…
A test is fully passed only if every run passed for that test.Wrong answer: 1Response Time (avg)98.48sResponse Time (max)98.48sResponse Time (total)98.48sA test is fully passed only if every run passed for that test.…
98.48sResponse Time (avg)…
261Total Input Tokens…
16Output Tokens…
60,256Reasoning Tokens…
Qwen3.8 27BCandidate responses were generated on local hardware. OpenRouter was used only for judging.
3.0Summarizes broad quality across our full private benchmark suite, so ranking reflects consistent performance.…
10.0Consistency score reflects run-to-run stability (10 = very consistent, even if consistently wrong).…
0.0%Attempt pass rate = passed attempts / total attempts across runs.…
0Flaky tests had mixed outcomes across runs (at least one pass and one fail).…
A test is fully passed only if every run passed for that test.Wrong answer: 1Response Time (avg)263.98sResponse Time (max)263.98sResponse Time (total)263.98sA test is fully passed only if every run passed for that test.…