The average score is effectively tied at 7.1 vs 7.1. MiMo-V2.5 (medium) has the lower benchmark cost at $0.130 vs ~$0.195. MiMo-V2.5 (medium) is faster at 49.20s vs 102.66s, with pass rates of 72.5% vs 69.6%.
Last updated at:
2026-10-05
Compared models
Candidate responses were generated on local hardware. OpenRouter was used only for judging. qwen3.8-flash-next-iq3_s · IQ3_S · Strata 0.1.39 · NVIDIA GeForce RTX 3090.
Rank
#154
Total Output Tokens
621,056
Response Time (avg)
102.66s
Total Cost
~$0.195Estimated GPU electricity cost only. It excludes the rest of the computer and hardware ownership. NVIDIA GeForce RTX 3090: 1.9676 h × 0.300 kW × €0.2896/kWh × 1.138 USD/EUR = ~$0.194534.
Rank
#157
Total Output Tokens
341,511
Response Time (avg)
49.20s
Total Cost
$0.130
Recommended modelMiMo-V2.5 (medium)
It has the best score here (7.1), while costing about 1.5x less than Qwen3.8 Flash Next (xhigh).
Detailed comparison
Metric
Qwen3.8 Flash NextQwen3.8 Flash NextxhighCandidate responses were generated on local hardware. OpenRouter was used only for judging. qwen3.8-flash-next-iq3_s · IQ3_S · Strata 0.1.39 · NVIDIA GeForce RTX 3090.Release: 2026-10-04
Qwen3.8 Flash NextQwen3.8 Flash NextxhighCandidate responses were generated on local hardware. OpenRouter was used only for judging. qwen3.8-flash-next-iq3_s · IQ3_S · Strata 0.1.39 · NVIDIA GeForce RTX 3090.Release: 2026-10-04
7.1Summarizes broad quality across our full private benchmark suite, so ranking reflects consistent performance.…
7.1Summarizes broad quality across our full private benchmark suite, so ranking reflects consistent performance.…
Rank
#154
#157
Reliability
9.9First-attempt success score: 10.0 means no retryable target API or rate-limit failures before successful calls; tracked failures lower the score.…
10.0First-attempt success score: 10.0 means no retryable target API or rate-limit failures before successful calls; tracked failures lower the score.…
Consistency
7.9Consistency score reflects run-to-run stability (10 = very consistent, even if consistently wrong).…
7.6Consistency score reflects run-to-run stability (10 = very consistent, even if consistently wrong).…
Attempts
69/69
69/69
Tests Correct
A test is fully passed only if every run passed for that test.Wrong answer: 6Did not follow instructions: 2Invalid tool call: 1Response Time (avg)102.66sResponse Time (max)674.08sResponse Time (total)2361.10sA test is fully passed only if every run passed for that test.…
72.5%Attempt pass rate = passed attempts / total attempts across runs.…
69.6%Attempt pass rate = passed attempts / total attempts across runs.…
Flaky tests
6Flaky tests had mixed outcomes across runs (at least one pass and one fail).…
7Flaky tests had mixed outcomes across runs (at least one pass and one fail).…
Total Runs
69Total Runs…
69Total Runs…
Cost per result
~1.390Estimated GPU electricity cost only. It excludes the rest of the computer and hardware ownership. NVIDIA GeForce RTX 3090: 1.9676 h × 0.300 kW × €0.2896/kWh × 1.138 USD/EUR = ~$0.194534.
3.491
Total Cost
~$0.195Estimated GPU electricity cost only. It excludes the rest of the computer and hardware ownership. NVIDIA GeForce RTX 3090: 1.9676 h × 0.300 kW × €0.2896/kWh × 1.138 USD/EUR = ~$0.194534.
$0.130
Input Price
N/AInput Price…
$0.140 / 1MInput Price…
Output Price
N/AOutput Price…
$0.280 / 1MOutput Price…
Cache Read Price
N/ACache Read Price…
$0.003 / 1MCache Read Price…
Cache Write Price
N/ACache Write Price…
N/ACache Write Price…
Total Input Tokens
301,749Total Input Tokens…
259,664Total Input Tokens…
Output Tokens
1,033Output Tokens…
8,414Output Tokens…
Reasoning Tokens
620,023Reasoning Tokens…
333,097Reasoning Tokens…
Response Time (avg)
102.66sResponse Time (avg)…
49.20sResponse Time (avg)…
Response Time (max)
674.08sResponse Time (max)…
316.27sResponse Time (max)…
Response Time (total)
2361.10sResponse Time (total)…
1131.69sResponse Time (total)…
Parameters
180B total (6B active)
310B total (15B active)
Availability
Weights available
Open source
Cache prices apply to input tokens. Reads reuse cached prompts; writes store them and can cost extra. Output tokens use the output price.
Model generation showcase
Hamster playing table tennis
Prompt: Create a detailed SVG illustration of a hamster playing table tennis.
#154 Qwen3.8 Flash Next
xhighCandidate responses were generated on local hardware. OpenRouter was used only for judging. qwen3.8-flash-next-iq3_s · IQ3_S · Strata 0.1.39 · NVIDIA GeForce RTX 3090.
Cost
~$0.010Estimated GPU electricity cost only. It excludes the rest of the computer and hardware ownership. NVIDIA GeForce RTX 3090: 0.0925 h × 0.300 kW × €0.2896/kWh × 1.138 USD/EUR = ~$0.009149.
Time
333.1s
Tokens
30,654 tok
#157 MiMo-V2.5
medium
Cost
$0.002
Time
54.8s
Tokens
5,247 tok
Rank
-
Cost
-
Time
-
Tokens
-
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
Qwen3.8 Flash NextCandidate responses were generated on local hardware. OpenRouter was used only for judging. qwen3.8-flash-next-iq3_s · IQ3_S · Strata 0.1.39 · NVIDIA GeForce RTX 3090.
4.7Summarizes broad quality across our full private benchmark suite, so ranking reflects consistent performance.…
3.1Consistency score reflects run-to-run stability (10 = very consistent, even if consistently wrong).…
33.3%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: 1Response Time (avg)173.66sResponse Time (max)173.66sResponse Time (total)173.66sA 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)113.57sResponse Time (max)113.57sResponse Time (total)113.57sA test is fully passed only if every run passed for that test.…
113.57sResponse Time (avg)…
154,208Total Input Tokens…
1,306Output Tokens…
13,157Reasoning Tokens…
Anti-AI Tricks
Score
Consistency
Attempt pass rate
Flaky tests
Tests Correct
Response Time (avg)
Input Tokens
Output Tokens
Reasoning Tokens
Qwen3.8 Flash NextCandidate responses were generated on local hardware. OpenRouter was used only for judging. qwen3.8-flash-next-iq3_s · IQ3_S · Strata 0.1.39 · NVIDIA GeForce RTX 3090.
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)6.73sResponse Time (max)18.55sResponse Time (total)26.91sA 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)4.14sResponse Time (max)12.41sResponse Time (total)16.57sA test is fully passed only if every run passed for that test.…
4.14sResponse Time (avg)…
621Total Input Tokens…
281Output Tokens…
1,739Reasoning Tokens…
Coding
Score
Consistency
Attempt pass rate
Flaky tests
Tests Correct
Response Time (avg)
Input Tokens
Output Tokens
Reasoning Tokens
Qwen3.8 Flash NextCandidate responses were generated on local hardware. OpenRouter was used only for judging. qwen3.8-flash-next-iq3_s · IQ3_S · Strata 0.1.39 · NVIDIA GeForce RTX 3090.
7.4Summarizes 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).…
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.Did not follow instructions: 1Response Time (avg)192.20sResponse Time (max)370.86sResponse Time (total)576.59sA test is fully passed only if every run passed for that test.…
6.2Summarizes broad quality across our full private benchmark suite, so ranking reflects consistent performance.…
4.7Consistency score reflects run-to-run stability (10 = very consistent, even if consistently wrong).…
66.7%Attempt pass rate = passed attempts / total attempts across runs.…
2Flaky 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: 2Response Time (avg)97.14sResponse Time (max)162.44sResponse Time (total)291.41sA test is fully passed only if every run passed for that test.…
97.14sResponse Time (avg)…
7,422Total Input Tokens…
557Output Tokens…
81,977Reasoning Tokens…
Combined
Score
Consistency
Attempt pass rate
Flaky tests
Tests Correct
Response Time (avg)
Input Tokens
Output Tokens
Reasoning Tokens
Qwen3.8 Flash NextCandidate responses were generated on local hardware. OpenRouter was used only for judging. qwen3.8-flash-next-iq3_s · IQ3_S · Strata 0.1.39 · NVIDIA GeForce RTX 3090.
7.3Summarizes broad quality across our full private benchmark suite, so ranking reflects consistent performance.…
5.8Consistency score reflects run-to-run stability (10 = very consistent, even if consistently wrong).…
83.3%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.Invalid tool call: 1Response Time (avg)92.24sResponse Time (max)143.77sResponse Time (total)184.49sA test is fully passed only if every run passed for that test.…
8.7Summarizes broad quality across our full private benchmark suite, so ranking reflects consistent performance.…
6.9Consistency score reflects run-to-run stability (10 = very consistent, even if consistently wrong).…
83.3%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.Invalid tool call: 1Response Time (avg)77.97sResponse Time (max)139.08sResponse Time (total)155.93sA test is fully passed only if every run passed for that test.…
77.97sResponse Time (avg)…
78,669Total Input Tokens…
4,656Output Tokens…
39,393Reasoning Tokens…
Data parsing and extraction
Score
Consistency
Attempt pass rate
Flaky tests
Tests Correct
Response Time (avg)
Input Tokens
Output Tokens
Reasoning Tokens
Qwen3.8 Flash NextCandidate responses were generated on local hardware. OpenRouter was used only for judging. qwen3.8-flash-next-iq3_s · IQ3_S · Strata 0.1.39 · NVIDIA GeForce RTX 3090.
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)9.24sResponse Time (max)9.63sResponse Time (total)18.48sA test is fully passed only if every run passed for that test.…
2.7Summarizes broad quality across our full private benchmark suite, so ranking reflects consistent performance.…
5.7Consistency score reflects run-to-run stability (10 = very consistent, even if consistently wrong).…
16.7%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.Extra formatting: 1Wrong answer: 1Response Time (avg)6.33sResponse Time (max)7.45sResponse Time (total)12.67sA test is fully passed only if every run passed for that test.…
6.33sResponse Time (avg)…
7,746Total Input Tokens…
306Output Tokens…
5,714Reasoning Tokens…
Domain specific
Score
Consistency
Attempt pass rate
Flaky tests
Tests Correct
Response Time (avg)
Input Tokens
Output Tokens
Reasoning Tokens
Qwen3.8 Flash NextCandidate responses were generated on local hardware. OpenRouter was used only for judging. qwen3.8-flash-next-iq3_s · IQ3_S · Strata 0.1.39 · NVIDIA GeForce RTX 3090.
3.5Summarizes broad quality across our full private benchmark suite, so ranking reflects consistent performance.…
4.4Consistency score reflects run-to-run stability (10 = very consistent, even if consistently wrong).…
33.3%Attempt pass rate = passed attempts / total attempts across runs.…
2Flaky 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)326.96sResponse Time (max)674.08sResponse Time (total)980.89sA 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.Extra formatting: 1No answer: 1Wrong answer: 1Response Time (avg)137.75sResponse Time (max)316.27sResponse Time (total)413.25sA test is fully passed only if every run passed for that test.…
137.75sResponse Time (avg)…
744Total Input Tokens…
495Output Tokens…
138,736Reasoning Tokens…
General Intelligence
Score
Consistency
Attempt pass rate
Flaky tests
Tests Correct
Response Time (avg)
Input Tokens
Output Tokens
Reasoning Tokens
Qwen3.8 Flash NextCandidate responses were generated on local hardware. OpenRouter was used only for judging. qwen3.8-flash-next-iq3_s · IQ3_S · Strata 0.1.39 · NVIDIA GeForce RTX 3090.
4.7Summarizes broad quality across our full private benchmark suite, so ranking reflects consistent performance.…
3.1Consistency score reflects run-to-run stability (10 = very consistent, even if consistently wrong).…
33.3%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)3.95sResponse Time (max)3.95sResponse Time (total)3.95sA test is fully passed only if every run passed for that test.…
5.4Summarizes broad quality across our full private benchmark suite, so ranking reflects consistent performance.…
2.5Consistency score reflects run-to-run stability (10 = very consistent, even if consistently wrong).…
66.7%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)5.37sResponse Time (max)5.37sResponse Time (total)5.37sA test is fully passed only if every run passed for that test.…
5.37sResponse Time (avg)…
492Total Input Tokens…
121Output Tokens…
418Reasoning Tokens…
Instructions following
Score
Consistency
Attempt pass rate
Flaky tests
Tests Correct
Response Time (avg)
Input Tokens
Output Tokens
Reasoning Tokens
Qwen3.8 Flash NextCandidate responses were generated on local hardware. OpenRouter was used only for judging. qwen3.8-flash-next-iq3_s · IQ3_S · Strata 0.1.39 · NVIDIA GeForce RTX 3090.
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)3.72sResponse Time (max)3.99sResponse Time (total)7.44sA test is fully passed only if every run passed for that test.…
9.9Summarizes 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)1.80sResponse Time (max)1.81sResponse Time (total)3.60sA test is fully passed only if every run passed for that test.…
1.80sResponse Time (avg)…
672Total Input Tokens…
88Output Tokens…
801Reasoning Tokens…
Puzzle Solving
Score
Consistency
Attempt pass rate
Flaky tests
Tests Correct
Response Time (avg)
Input Tokens
Output Tokens
Reasoning Tokens
Qwen3.8 Flash NextCandidate responses were generated on local hardware. OpenRouter was used only for judging. qwen3.8-flash-next-iq3_s · IQ3_S · Strata 0.1.39 · NVIDIA GeForce RTX 3090.
7.7Summarizes 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).…
66.7%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)19.51sResponse Time (max)39.77sResponse Time (total)58.54sA test is fully passed only if every run passed for that test.…
8.2Summarizes 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).…
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.No answer: 1Response Time (avg)20.25sResponse Time (max)57.93sResponse Time (total)60.76sA test is fully passed only if every run passed for that test.…
20.25sResponse Time (avg)…
660Total Input Tokens…
279Output Tokens…
33,254Reasoning Tokens…
Tool Calling
Score
Consistency
Attempt pass rate
Flaky tests
Tests Correct
Response Time (avg)
Input Tokens
Output Tokens
Reasoning Tokens
Qwen3.8 Flash NextCandidate responses were generated on local hardware. OpenRouter was used only for judging. qwen3.8-flash-next-iq3_s · IQ3_S · Strata 0.1.39 · NVIDIA GeForce RTX 3090.
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)5.11sResponse Time (max)5.11sResponse Time (total)5.11sA 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)7.29sResponse Time (max)7.29sResponse Time (total)7.29sA test is fully passed only if every run passed for that test.…
7.29sResponse Time (avg)…
8,220Total Input Tokens…
303Output Tokens…
2,424Reasoning Tokens…
Trivia
Score
Consistency
Attempt pass rate
Flaky tests
Tests Correct
Response Time (avg)
Input Tokens
Output Tokens
Reasoning Tokens
Qwen3.8 Flash NextCandidate responses were generated on local hardware. OpenRouter was used only for judging. qwen3.8-flash-next-iq3_s · IQ3_S · Strata 0.1.39 · NVIDIA GeForce RTX 3090.
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)325.04sResponse Time (max)325.04sResponse Time (total)325.04sA 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)51.29sResponse Time (max)51.29sResponse Time (total)51.29sA test is fully passed only if every run passed for that test.…