Qwen: Qwen3.8 2.4T A95B

Qwen3.8 2.4T A95B is an open-weight sparse mixture-of-experts model from Qwen and the open-weight variant of [Qwen3.8 Max](/qwen/qwen3.8-max), with 95 billion active parameters out of 2.4 trillion to…

qwen/qwen3.8-2.4t-a95b-20260812

Qwen3.8 2.4T A95B is an open-weight sparse mixture-of-experts model from Qwen and the open-weight variant of Qwen3.8 Max, with 95 billion active parameters out of 2.4 trillion total. It is suited for coding, research, complex reasoning, and agentic workflows.

Specifications & capabilities

Input
text
Output
text
Context window · tokens
1048576
Published model ID
Qwen3.8-2.4T-A95B
Capabilities
multimodal, promptCaching, reasoning, structuredOutputs, toolCalling, visionInput
Description source ↗

Evaluations

1–19 / 19

OpenRouter Benchmarks

91.771%

source id
openRouterBenchmarks
leaderboard name
GPQA Diamond · Accuracy · Venice
dimension
knowledge
metric
GPQA Diamond · Accuracy
rank
1
sample size
2147
aggregation scope
status
published
harness
url
https://github.com/OpenRouterTeam/benchmark-harness
name
OpenRouter benchmark harness
direction
higherIsBetter
run_count
11
conditions
window
newest 90 days; successful visible nonzero runs
dataset
benchmark-leaderboard
benchmark
id
gpqa_diamond
name
GPQA Diamond
record_kind
aggregate
provider_name
Venice
attribution
Source: OpenRouter (openrouter.ai), licensed under CC BY 4.0: reuse and republish with attribution to OpenRouter.
last_run_at
2026-10-31T08:34:22.445Z
license_url
https://creativecommons.org/licenses/by/4.0/
methodology
Rows aggregate the benchmark-runs dataset exactly as the leaderboard pages do, over each benchmark's non-zero-score runs from the newest 90 days: accuracy is the task-weighted mean of per-run accuracy, accuracy_std_dev is the population standard deviation of per-run accuracy (null with a single run), and the per-task averages divide summed cost, generation time, and output tokens by summed tasks. Rows are ordered by accuracy descending, then model name.
metric_kind
GPQA Diamond · Accuracy
rank_method
competition rank among reported evaluations in identical conditions, including unlinked observations
sample_size
2147
source_file
benchmark-leaderboard.json
source_page
https://openrouter.ai/benchmarks/gpqa-diamond
measurements
accuracy
0.9177144177922683
accuracy_std_dev
0.013751154902061724
cost_per_task_usd
0.2460820894271076
output_tokens_per_task
37154.30647414998
latency_seconds_per_task
409.07095482068
published_at
2026-11-16T01:46:02.653Z
snapshot_date
2026-11-16
source_revision
7ccaf17664252a3053dcb7df8ad0a46e72954a3b4524ddc1c8a7ef8a41b15568
transformations
TensorTrace preserves source results, converts latency to seconds and derives ranks within matching conditions. Individual runs have no derived rank.
metric_population
39
benchmark_evaluation
true
openrouter_permaslug
qwen/qwen3.8-2.4t-a95b-20260812
source model id
qwen/qwen3.8-2.4t-a95b-20260812
Original result ↗

OpenRouter Benchmarks

74.722%

source id
openRouterBenchmarks
leaderboard name
tau2-bench Verified Airline · Accuracy · Modal
dimension
agentic
metric
tau2-bench Verified Airline · Accuracy
rank
5
sample size
300
aggregation scope
status
published
harness
url
https://github.com/OpenRouterTeam/benchmark-harness
name
OpenRouter benchmark harness
direction
higherIsBetter
run_count
6
conditions
window
newest 90 days; successful visible nonzero runs
dataset
benchmark-leaderboard
benchmark
id
tau_bench_verified_airline
name
tau2-bench Verified Airline
record_kind
aggregate
provider_name
Modal
attribution
Source: OpenRouter (openrouter.ai), licensed under CC BY 4.0: reuse and republish with attribution to OpenRouter.
last_run_at
2026-10-10T11:19:48.441Z
license_url
https://creativecommons.org/licenses/by/4.0/
methodology
Rows aggregate the benchmark-runs dataset exactly as the leaderboard pages do, over each benchmark's non-zero-score runs from the newest 90 days: accuracy is the task-weighted mean of per-run accuracy, accuracy_std_dev is the population standard deviation of per-run accuracy (null with a single run), and the per-task averages divide summed cost, generation time, and output tokens by summed tasks. Rows are ordered by accuracy descending, then model name.
metric_kind
tau2-bench Verified Airline · Accuracy
rank_method
competition rank among reported evaluations in identical conditions, including unlinked observations
sample_size
300
source_file
benchmark-leaderboard.json
source_page
https://openrouter.ai/benchmarks/tau2-bench-airline
measurements
accuracy
0.7472221666666666
accuracy_std_dev
0.011613686586906374
cost_per_task_usd
0.17439429333333334
output_tokens_per_task
12499.9
latency_seconds_per_task
161.42059666666668
published_at
2026-11-16T01:46:02.653Z
snapshot_date
2026-11-16
source_revision
7ccaf17664252a3053dcb7df8ad0a46e72954a3b4524ddc1c8a7ef8a41b15568
transformations
TensorTrace preserves source results, converts latency to seconds and derives ranks within matching conditions. Individual runs have no derived rank.
metric_population
11
benchmark_evaluation
true
openrouter_permaslug
qwen/qwen3.8-2.4t-a95b-20260812
source model id
qwen/qwen3.8-2.4t-a95b-20260812
Original result ↗

OpenRouter Benchmarks

77.429%

source id
openRouterBenchmarks
leaderboard name
tau2-bench Verified Airline · Accuracy · DeepInfra
dimension
agentic
metric
tau2-bench Verified Airline · Accuracy
rank
6
sample size
350
aggregation scope
status
published
harness
url
https://github.com/OpenRouterTeam/benchmark-harness
name
OpenRouter benchmark harness
direction
higherIsBetter
run_count
7
conditions
window
newest 90 days; successful visible nonzero runs
dataset
benchmark-leaderboard
benchmark
id
tau_bench_verified_airline
name
tau2-bench Verified Airline
record_kind
aggregate
provider_name
DeepInfra
attribution
Source: OpenRouter (openrouter.ai), licensed under CC BY 4.0: reuse and republish with attribution to OpenRouter.
last_run_at
2026-10-10T11:06:43.832Z
license_url
https://creativecommons.org/licenses/by/4.0/
methodology
Rows aggregate the benchmark-runs dataset exactly as the leaderboard pages do, over each benchmark's non-zero-score runs from the newest 90 days: accuracy is the task-weighted mean of per-run accuracy, accuracy_std_dev is the population standard deviation of per-run accuracy (null with a single run), and the per-task averages divide summed cost, generation time, and output tokens by summed tasks. Rows are ordered by accuracy descending, then model name.
metric_kind
tau2-bench Verified Airline · Accuracy
rank_method
competition rank among reported evaluations in identical conditions, including unlinked observations
sample_size
350
source_file
benchmark-leaderboard.json
source_page
https://openrouter.ai/benchmarks/tau2-bench-airline
measurements
accuracy
0.7742857142857142
accuracy_std_dev
0.026945723560596418
cost_per_task_usd
0.21623159142857146
output_tokens_per_task
12602.945714285714
latency_seconds_per_task
562.3065571428571
published_at
2026-11-16T01:46:02.653Z
snapshot_date
2026-11-16
source_revision
7ccaf17664252a3053dcb7df8ad0a46e72954a3b4524ddc1c8a7ef8a41b15568
transformations
TensorTrace preserves source results, converts latency to seconds and derives ranks within matching conditions. Individual runs have no derived rank.
metric_population
63
benchmark_evaluation
true
openrouter_permaslug
qwen/qwen3.8-2.4t-a95b-20260812
source model id
qwen/qwen3.8-2.4t-a95b-20260812
Original result ↗

OpenRouter Benchmarks

86.821%

source id
openRouterBenchmarks
leaderboard name
GPQA Diamond · Accuracy · Modal
dimension
knowledge
metric
GPQA Diamond · Accuracy
rank
7
sample size
1386
aggregation scope
status
published
harness
url
https://github.com/OpenRouterTeam/benchmark-harness
name
OpenRouter benchmark harness
direction
higherIsBetter
run_count
7
conditions
window
newest 90 days; successful visible nonzero runs
dataset
benchmark-leaderboard
benchmark
id
gpqa_diamond
name
GPQA Diamond
record_kind
aggregate
provider_name
Modal
attribution
Source: OpenRouter (openrouter.ai), licensed under CC BY 4.0: reuse and republish with attribution to OpenRouter.
last_run_at
2026-10-10T08:21:46.324Z
license_url
https://creativecommons.org/licenses/by/4.0/
methodology
Rows aggregate the benchmark-runs dataset exactly as the leaderboard pages do, over each benchmark's non-zero-score runs from the newest 90 days: accuracy is the task-weighted mean of per-run accuracy, accuracy_std_dev is the population standard deviation of per-run accuracy (null with a single run), and the per-task averages divide summed cost, generation time, and output tokens by summed tasks. Rows are ordered by accuracy descending, then model name.
metric_kind
GPQA Diamond · Accuracy
rank_method
competition rank among reported evaluations in identical conditions, including unlinked observations
sample_size
1386
source_file
benchmark-leaderboard.json
source_page
https://openrouter.ai/benchmarks/gpqa-diamond
measurements
accuracy
0.8682058571428571
accuracy_std_dev
0.007673379706651383
cost_per_task_usd
0.07688844300144301
output_tokens_per_task
12574.14357864358
latency_seconds_per_task
83.58213492063493
published_at
2026-11-16T01:46:02.653Z
snapshot_date
2026-11-16
source_revision
7ccaf17664252a3053dcb7df8ad0a46e72954a3b4524ddc1c8a7ef8a41b15568
transformations
TensorTrace preserves source results, converts latency to seconds and derives ranks within matching conditions. Individual runs have no derived rank.
metric_population
11
benchmark_evaluation
true
openrouter_permaslug
qwen/qwen3.8-2.4t-a95b-20260812
source model id
qwen/qwen3.8-2.4t-a95b-20260812
Original result ↗

OpenRouter Benchmarks

75.879%

source id
openRouterBenchmarks
leaderboard name
tau2-bench Verified Airline · Accuracy · Alibaba
dimension
agentic
metric
tau2-bench Verified Airline · Accuracy
rank
8
sample size
550
aggregation scope
status
published
harness
url
https://github.com/OpenRouterTeam/benchmark-harness
name
OpenRouter benchmark harness
direction
higherIsBetter
run_count
11
conditions
window
newest 90 days; successful visible nonzero runs
dataset
benchmark-leaderboard
benchmark
id
tau_bench_verified_airline
name
tau2-bench Verified Airline
record_kind
aggregate
provider_name
Alibaba
attribution
Source: OpenRouter (openrouter.ai), licensed under CC BY 4.0: reuse and republish with attribution to OpenRouter.
last_run_at
2026-10-31T11:13:27.901Z
license_url
https://creativecommons.org/licenses/by/4.0/
methodology
Rows aggregate the benchmark-runs dataset exactly as the leaderboard pages do, over each benchmark's non-zero-score runs from the newest 90 days: accuracy is the task-weighted mean of per-run accuracy, accuracy_std_dev is the population standard deviation of per-run accuracy (null with a single run), and the per-task averages divide summed cost, generation time, and output tokens by summed tasks. Rows are ordered by accuracy descending, then model name.
metric_kind
tau2-bench Verified Airline · Accuracy
rank_method
competition rank among reported evaluations in identical conditions, including unlinked observations
sample_size
550
source_file
benchmark-leaderboard.json
source_page
https://openrouter.ai/benchmarks/tau2-bench-airline
measurements
accuracy
0.7587878181818182
accuracy_std_dev
0.01653114932933463
cost_per_task_usd
0.17482494545454544
output_tokens_per_task
10740.534545454546
latency_seconds_per_task
315.7564018181818
published_at
2026-11-16T01:46:02.653Z
snapshot_date
2026-11-16
source_revision
7ccaf17664252a3053dcb7df8ad0a46e72954a3b4524ddc1c8a7ef8a41b15568
transformations
TensorTrace preserves source results, converts latency to seconds and derives ranks within matching conditions. Individual runs have no derived rank.
metric_population
29
benchmark_evaluation
true
openrouter_permaslug
qwen/qwen3.8-2.4t-a95b-20260812
source model id
qwen/qwen3.8-2.4t-a95b-20260812
Original result ↗

OpenRouter Benchmarks

74.519%

source id
openRouterBenchmarks
leaderboard name
tau2-bench Verified Airline · Accuracy · Together
dimension
agentic
metric
tau2-bench Verified Airline · Accuracy
rank
8
sample size
450
aggregation scope
status
published
harness
url
https://github.com/OpenRouterTeam/benchmark-harness
name
OpenRouter benchmark harness
direction
higherIsBetter
run_count
9
conditions
window
newest 90 days; successful visible nonzero runs
dataset
benchmark-leaderboard
benchmark
id
tau_bench_verified_airline
name
tau2-bench Verified Airline
record_kind
aggregate
provider_name
Together
attribution
Source: OpenRouter (openrouter.ai), licensed under CC BY 4.0: reuse and republish with attribution to OpenRouter.
last_run_at
2026-10-31T11:21:59.486Z
license_url
https://creativecommons.org/licenses/by/4.0/
methodology
Rows aggregate the benchmark-runs dataset exactly as the leaderboard pages do, over each benchmark's non-zero-score runs from the newest 90 days: accuracy is the task-weighted mean of per-run accuracy, accuracy_std_dev is the population standard deviation of per-run accuracy (null with a single run), and the per-task averages divide summed cost, generation time, and output tokens by summed tasks. Rows are ordered by accuracy descending, then model name.
metric_kind
tau2-bench Verified Airline · Accuracy
rank_method
competition rank among reported evaluations in identical conditions, including unlinked observations
sample_size
450
source_file
benchmark-leaderboard.json
source_page
https://openrouter.ai/benchmarks/tau2-bench-airline
measurements
accuracy
0.7451852222222222
accuracy_std_dev
0.017716056438394953
cost_per_task_usd
0.21677278888888887
output_tokens_per_task
11557.691111111111
latency_seconds_per_task
284.79118666666665
published_at
2026-11-16T01:46:02.653Z
snapshot_date
2026-11-16
source_revision
7ccaf17664252a3053dcb7df8ad0a46e72954a3b4524ddc1c8a7ef8a41b15568
transformations
TensorTrace preserves source results, converts latency to seconds and derives ranks within matching conditions. Individual runs have no derived rank.
metric_population
22
benchmark_evaluation
true
openrouter_permaslug
qwen/qwen3.8-2.4t-a95b-20260812
source model id
qwen/qwen3.8-2.4t-a95b-20260812
Original result ↗

OpenRouter Benchmarks

74.267%

source id
openRouterBenchmarks
leaderboard name
tau2-bench Verified Airline · Accuracy · SiliconFlow
dimension
agentic
metric
tau2-bench Verified Airline · Accuracy
rank
10
sample size
500
aggregation scope
status
published
harness
url
https://github.com/OpenRouterTeam/benchmark-harness
name
OpenRouter benchmark harness
direction
higherIsBetter
run_count
10
conditions
window
newest 90 days; successful visible nonzero runs
dataset
benchmark-leaderboard
benchmark
id
tau_bench_verified_airline
name
tau2-bench Verified Airline
record_kind
aggregate
provider_name
SiliconFlow
attribution
Source: OpenRouter (openrouter.ai), licensed under CC BY 4.0: reuse and republish with attribution to OpenRouter.
last_run_at
2026-10-31T11:21:43.861Z
license_url
https://creativecommons.org/licenses/by/4.0/
methodology
Rows aggregate the benchmark-runs dataset exactly as the leaderboard pages do, over each benchmark's non-zero-score runs from the newest 90 days: accuracy is the task-weighted mean of per-run accuracy, accuracy_std_dev is the population standard deviation of per-run accuracy (null with a single run), and the per-task averages divide summed cost, generation time, and output tokens by summed tasks. Rows are ordered by accuracy descending, then model name.
metric_kind
tau2-bench Verified Airline · Accuracy
rank_method
competition rank among reported evaluations in identical conditions, including unlinked observations
sample_size
500
source_file
benchmark-leaderboard.json
source_page
https://openrouter.ai/benchmarks/tau2-bench-airline
measurements
accuracy
0.7426666000000001
accuracy_std_dev
0.04434197309367279
cost_per_task_usd
0.16675142799999998
output_tokens_per_task
9682.606
latency_seconds_per_task
302.04735600000004
published_at
2026-11-16T01:46:02.653Z
snapshot_date
2026-11-16
source_revision
7ccaf17664252a3053dcb7df8ad0a46e72954a3b4524ddc1c8a7ef8a41b15568
transformations
TensorTrace preserves source results, converts latency to seconds and derives ranks within matching conditions. Individual runs have no derived rank.
metric_population
28
benchmark_evaluation
true
openrouter_permaslug
qwen/qwen3.8-2.4t-a95b-20260812
source model id
qwen/qwen3.8-2.4t-a95b-20260812
Original result ↗

OpenRouter Benchmarks

86.308%

source id
openRouterBenchmarks
leaderboard name
GPQA Diamond · Accuracy · Alibaba
dimension
knowledge
metric
GPQA Diamond · Accuracy
rank
11
sample size
2376
aggregation scope
status
published
harness
url
https://github.com/OpenRouterTeam/benchmark-harness
name
OpenRouter benchmark harness
direction
higherIsBetter
run_count
12
conditions
window
newest 90 days; successful visible nonzero runs
dataset
benchmark-leaderboard
benchmark
id
gpqa_diamond
name
GPQA Diamond
record_kind
aggregate
provider_name
Alibaba
attribution
Source: OpenRouter (openrouter.ai), licensed under CC BY 4.0: reuse and republish with attribution to OpenRouter.
last_run_at
2026-10-31T08:13:57.551Z
license_url
https://creativecommons.org/licenses/by/4.0/
methodology
Rows aggregate the benchmark-runs dataset exactly as the leaderboard pages do, over each benchmark's non-zero-score runs from the newest 90 days: accuracy is the task-weighted mean of per-run accuracy, accuracy_std_dev is the population standard deviation of per-run accuracy (null with a single run), and the per-task averages divide summed cost, generation time, and output tokens by summed tasks. Rows are ordered by accuracy descending, then model name.
metric_kind
GPQA Diamond · Accuracy
rank_method
competition rank among reported evaluations in identical conditions, including unlinked observations
sample_size
2376
source_file
benchmark-leaderboard.json
source_page
https://openrouter.ai/benchmarks/gpqa-diamond
measurements
accuracy
0.8630751666666665
accuracy_std_dev
0.009048483766478349
cost_per_task_usd
0.0646409191919192
output_tokens_per_task
10505.142255892257
latency_seconds_per_task
213.6551308922559
published_at
2026-11-16T01:46:02.653Z
snapshot_date
2026-11-16
source_revision
7ccaf17664252a3053dcb7df8ad0a46e72954a3b4524ddc1c8a7ef8a41b15568
transformations
TensorTrace preserves source results, converts latency to seconds and derives ranks within matching conditions. Individual runs have no derived rank.
metric_population
29
benchmark_evaluation
true
openrouter_permaslug
qwen/qwen3.8-2.4t-a95b-20260812
source model id
qwen/qwen3.8-2.4t-a95b-20260812
Original result ↗

OpenRouter Benchmarks

86.241%

source id
openRouterBenchmarks
leaderboard name
GPQA Diamond · Accuracy · SiliconFlow
dimension
knowledge
metric
GPQA Diamond · Accuracy
rank
12
sample size
2178
aggregation scope
status
published
harness
url
https://github.com/OpenRouterTeam/benchmark-harness
name
OpenRouter benchmark harness
direction
higherIsBetter
run_count
11
conditions
window
newest 90 days; successful visible nonzero runs
dataset
benchmark-leaderboard
benchmark
id
gpqa_diamond
name
GPQA Diamond
record_kind
aggregate
provider_name
SiliconFlow
attribution
Source: OpenRouter (openrouter.ai), licensed under CC BY 4.0: reuse and republish with attribution to OpenRouter.
last_run_at
2026-10-31T08:01:54.465Z
license_url
https://creativecommons.org/licenses/by/4.0/
methodology
Rows aggregate the benchmark-runs dataset exactly as the leaderboard pages do, over each benchmark's non-zero-score runs from the newest 90 days: accuracy is the task-weighted mean of per-run accuracy, accuracy_std_dev is the population standard deviation of per-run accuracy (null with a single run), and the per-task averages divide summed cost, generation time, and output tokens by summed tasks. Rows are ordered by accuracy descending, then model name.
metric_kind
GPQA Diamond · Accuracy
rank_method
competition rank among reported evaluations in identical conditions, including unlinked observations
sample_size
2178
source_file
benchmark-leaderboard.json
source_page
https://openrouter.ai/benchmarks/gpqa-diamond
measurements
accuracy
0.8624118181818182
accuracy_std_dev
0.010959488075454655
cost_per_task_usd
0.0702179696969697
output_tokens_per_task
11433.751147842057
latency_seconds_per_task
231.2111772268136
published_at
2026-11-16T01:46:02.653Z
snapshot_date
2026-11-16
source_revision
7ccaf17664252a3053dcb7df8ad0a46e72954a3b4524ddc1c8a7ef8a41b15568
transformations
TensorTrace preserves source results, converts latency to seconds and derives ranks within matching conditions. Individual runs have no derived rank.
metric_population
34
benchmark_evaluation
true
openrouter_permaslug
qwen/qwen3.8-2.4t-a95b-20260812
source model id
qwen/qwen3.8-2.4t-a95b-20260812
Original result ↗

OpenRouter Benchmarks

74.533%

source id
openRouterBenchmarks
leaderboard name
tau2-bench Verified Airline · Accuracy · Novita
dimension
agentic
metric
tau2-bench Verified Airline · Accuracy
rank
14
sample size
500
aggregation scope
status
published
harness
url
https://github.com/OpenRouterTeam/benchmark-harness
name
OpenRouter benchmark harness
direction
higherIsBetter
run_count
10
conditions
window
newest 90 days; successful visible nonzero runs
dataset
benchmark-leaderboard
benchmark
id
tau_bench_verified_airline
name
tau2-bench Verified Airline
record_kind
aggregate
provider_name
Novita
attribution
Source: OpenRouter (openrouter.ai), licensed under CC BY 4.0: reuse and republish with attribution to OpenRouter.
last_run_at
2026-10-31T12:25:36.323Z
license_url
https://creativecommons.org/licenses/by/4.0/
methodology
Rows aggregate the benchmark-runs dataset exactly as the leaderboard pages do, over each benchmark's non-zero-score runs from the newest 90 days: accuracy is the task-weighted mean of per-run accuracy, accuracy_std_dev is the population standard deviation of per-run accuracy (null with a single run), and the per-task averages divide summed cost, generation time, and output tokens by summed tasks. Rows are ordered by accuracy descending, then model name.
metric_kind
tau2-bench Verified Airline · Accuracy
rank_method
competition rank among reported evaluations in identical conditions, including unlinked observations
sample_size
500
source_file
benchmark-leaderboard.json
source_page
https://openrouter.ai/benchmarks/tau2-bench-airline
measurements
accuracy
0.7453333
accuracy_std_dev
0.02454934079379731
cost_per_task_usd
0.17119798400000003
output_tokens_per_task
11023.856
latency_seconds_per_task
320.38772600000004
published_at
2026-11-16T01:46:02.653Z
snapshot_date
2026-11-16
source_revision
7ccaf17664252a3053dcb7df8ad0a46e72954a3b4524ddc1c8a7ef8a41b15568
transformations
TensorTrace preserves source results, converts latency to seconds and derives ranks within matching conditions. Individual runs have no derived rank.
metric_population
52
benchmark_evaluation
true
openrouter_permaslug
qwen/qwen3.8-2.4t-a95b-20260812
source model id
qwen/qwen3.8-2.4t-a95b-20260812
Original result ↗

OpenRouter Benchmarks

87.149%

source id
openRouterBenchmarks
leaderboard name
GPQA Diamond · Accuracy · DeepInfra
dimension
knowledge
metric
GPQA Diamond · Accuracy
rank
15
sample size
1577
aggregation scope
status
published
harness
url
https://github.com/OpenRouterTeam/benchmark-harness
name
OpenRouter benchmark harness
direction
higherIsBetter
run_count
8
conditions
window
newest 90 days; successful visible nonzero runs
dataset
benchmark-leaderboard
benchmark
id
gpqa_diamond
name
GPQA Diamond
record_kind
aggregate
provider_name
DeepInfra
attribution
Source: OpenRouter (openrouter.ai), licensed under CC BY 4.0: reuse and republish with attribution to OpenRouter.
last_run_at
2026-10-10T08:22:14.215Z
license_url
https://creativecommons.org/licenses/by/4.0/
methodology
Rows aggregate the benchmark-runs dataset exactly as the leaderboard pages do, over each benchmark's non-zero-score runs from the newest 90 days: accuracy is the task-weighted mean of per-run accuracy, accuracy_std_dev is the population standard deviation of per-run accuracy (null with a single run), and the per-task averages divide summed cost, generation time, and output tokens by summed tasks. Rows are ordered by accuracy descending, then model name.
metric_kind
GPQA Diamond · Accuracy
rank_method
competition rank among reported evaluations in identical conditions, including unlinked observations
sample_size
1577
source_file
benchmark-leaderboard.json
source_page
https://openrouter.ai/benchmarks/gpqa-diamond
measurements
accuracy
0.8714859264426126
accuracy_std_dev
0.00752103815972237
cost_per_task_usd
0.06945589727330374
output_tokens_per_task
11308.629676601142
latency_seconds_per_task
276.8587152821813
published_at
2026-11-16T01:46:02.653Z
snapshot_date
2026-11-16
source_revision
7ccaf17664252a3053dcb7df8ad0a46e72954a3b4524ddc1c8a7ef8a41b15568
transformations
TensorTrace preserves source results, converts latency to seconds and derives ranks within matching conditions. Individual runs have no derived rank.
metric_population
71
benchmark_evaluation
true
openrouter_permaslug
qwen/qwen3.8-2.4t-a95b-20260812
source model id
qwen/qwen3.8-2.4t-a95b-20260812
Original result ↗

OpenRouter Benchmarks

87.003%

source id
openRouterBenchmarks
leaderboard name
GPQA Diamond · Accuracy · Together
dimension
knowledge
metric
GPQA Diamond · Accuracy
rank
15
sample size
1980
aggregation scope
status
published
harness
url
https://github.com/OpenRouterTeam/benchmark-harness
name
OpenRouter benchmark harness
direction
higherIsBetter
run_count
10
conditions
window
newest 90 days; successful visible nonzero runs
dataset
benchmark-leaderboard
benchmark
id
gpqa_diamond
name
GPQA Diamond
record_kind
aggregate
provider_name
Together
attribution
Source: OpenRouter (openrouter.ai), licensed under CC BY 4.0: reuse and republish with attribution to OpenRouter.
last_run_at
2026-10-31T08:03:09.289Z
license_url
https://creativecommons.org/licenses/by/4.0/
methodology
Rows aggregate the benchmark-runs dataset exactly as the leaderboard pages do, over each benchmark's non-zero-score runs from the newest 90 days: accuracy is the task-weighted mean of per-run accuracy, accuracy_std_dev is the population standard deviation of per-run accuracy (null with a single run), and the per-task averages divide summed cost, generation time, and output tokens by summed tasks. Rows are ordered by accuracy descending, then model name.
metric_kind
GPQA Diamond · Accuracy
rank_method
competition rank among reported evaluations in identical conditions, including unlinked observations
sample_size
1980
source_file
benchmark-leaderboard.json
source_page
https://openrouter.ai/benchmarks/gpqa-diamond
measurements
accuracy
0.8700336
accuracy_std_dev
0.008343089106559986
cost_per_task_usd
0.07480730858585859
output_tokens_per_task
12161.163636363637
latency_seconds_per_task
106.90641464646465
published_at
2026-11-16T01:46:02.653Z
snapshot_date
2026-11-16
source_revision
7ccaf17664252a3053dcb7df8ad0a46e72954a3b4524ddc1c8a7ef8a41b15568
transformations
TensorTrace preserves source results, converts latency to seconds and derives ranks within matching conditions. Individual runs have no derived rank.
metric_population
25
benchmark_evaluation
true
openrouter_permaslug
qwen/qwen3.8-2.4t-a95b-20260812
source model id
qwen/qwen3.8-2.4t-a95b-20260812
Original result ↗

OpenRouter Benchmarks

86.762%

source id
openRouterBenchmarks
leaderboard name
GPQA Diamond · Accuracy · Novita
dimension
knowledge
metric
GPQA Diamond · Accuracy
rank
16
sample size
2178
aggregation scope
status
published
harness
url
https://github.com/OpenRouterTeam/benchmark-harness
name
OpenRouter benchmark harness
direction
higherIsBetter
run_count
11
conditions
window
newest 90 days; successful visible nonzero runs
dataset
benchmark-leaderboard
benchmark
id
gpqa_diamond
name
GPQA Diamond
record_kind
aggregate
provider_name
Novita
attribution
Source: OpenRouter (openrouter.ai), licensed under CC BY 4.0: reuse and republish with attribution to OpenRouter.
last_run_at
2026-10-31T08:17:29.729Z
license_url
https://creativecommons.org/licenses/by/4.0/
methodology
Rows aggregate the benchmark-runs dataset exactly as the leaderboard pages do, over each benchmark's non-zero-score runs from the newest 90 days: accuracy is the task-weighted mean of per-run accuracy, accuracy_std_dev is the population standard deviation of per-run accuracy (null with a single run), and the per-task averages divide summed cost, generation time, and output tokens by summed tasks. Rows are ordered by accuracy descending, then model name.
metric_kind
GPQA Diamond · Accuracy
rank_method
competition rank among reported evaluations in identical conditions, including unlinked observations
sample_size
2178
source_file
benchmark-leaderboard.json
source_page
https://openrouter.ai/benchmarks/gpqa-diamond
measurements
accuracy
0.8676153636363637
accuracy_std_dev
0.004014278793432891
cost_per_task_usd
0.0702307419651056
output_tokens_per_task
11437.757575757576
latency_seconds_per_task
231.76075160697889
published_at
2026-11-16T01:46:02.653Z
snapshot_date
2026-11-16
source_revision
7ccaf17664252a3053dcb7df8ad0a46e72954a3b4524ddc1c8a7ef8a41b15568
transformations
TensorTrace preserves source results, converts latency to seconds and derives ranks within matching conditions. Individual runs have no derived rank.
metric_population
59
benchmark_evaluation
true
openrouter_permaslug
qwen/qwen3.8-2.4t-a95b-20260812
source model id
qwen/qwen3.8-2.4t-a95b-20260812
Original result ↗

OpenRouter Benchmarks

72.4%

source id
openRouterBenchmarks
leaderboard name
tau2-bench Verified Airline · Accuracy · Venice
dimension
agentic
metric
tau2-bench Verified Airline · Accuracy
rank
18
sample size
500
aggregation scope
status
published
harness
url
https://github.com/OpenRouterTeam/benchmark-harness
name
OpenRouter benchmark harness
direction
higherIsBetter
run_count
10
conditions
window
newest 90 days; successful visible nonzero runs
dataset
benchmark-leaderboard
benchmark
id
tau_bench_verified_airline
name
tau2-bench Verified Airline
record_kind
aggregate
provider_name
Venice
attribution
Source: OpenRouter (openrouter.ai), licensed under CC BY 4.0: reuse and republish with attribution to OpenRouter.
last_run_at
2026-10-31T12:56:09.161Z
license_url
https://creativecommons.org/licenses/by/4.0/
methodology
Rows aggregate the benchmark-runs dataset exactly as the leaderboard pages do, over each benchmark's non-zero-score runs from the newest 90 days: accuracy is the task-weighted mean of per-run accuracy, accuracy_std_dev is the population standard deviation of per-run accuracy (null with a single run), and the per-task averages divide summed cost, generation time, and output tokens by summed tasks. Rows are ordered by accuracy descending, then model name.
metric_kind
tau2-bench Verified Airline · Accuracy
rank_method
competition rank among reported evaluations in identical conditions, including unlinked observations
sample_size
500
source_file
benchmark-leaderboard.json
source_page
https://openrouter.ai/benchmarks/tau2-bench-airline
measurements
accuracy
0.724
accuracy_std_dev
0.02351363670723862
cost_per_task_usd
0.26574487199999997
output_tokens_per_task
11402.85
latency_seconds_per_task
177.43486199999998
published_at
2026-11-16T01:46:02.653Z
snapshot_date
2026-11-16
source_revision
7ccaf17664252a3053dcb7df8ad0a46e72954a3b4524ddc1c8a7ef8a41b15568
transformations
TensorTrace preserves source results, converts latency to seconds and derives ranks within matching conditions. Individual runs have no derived rank.
metric_population
38
benchmark_evaluation
true
openrouter_permaslug
qwen/qwen3.8-2.4t-a95b-20260812
source model id
qwen/qwen3.8-2.4t-a95b-20260812
Original result ↗

OpenRouter Benchmarks

74.815%

source id
openRouterBenchmarks
leaderboard name
tau2-bench Verified Airline · Accuracy · auto-routing
dimension
agentic
metric
tau2-bench Verified Airline · Accuracy
rank
33
sample size
450
aggregation scope
status
published
harness
url
https://github.com/OpenRouterTeam/benchmark-harness
name
OpenRouter benchmark harness
direction
higherIsBetter
run_count
9
conditions
window
newest 90 days; successful visible nonzero runs
dataset
benchmark-leaderboard
benchmark
id
tau_bench_verified_airline
name
tau2-bench Verified Airline
record_kind
aggregate
provider_name
auto-routing
attribution
Source: OpenRouter (openrouter.ai), licensed under CC BY 4.0: reuse and republish with attribution to OpenRouter.
last_run_at
2026-10-31T12:37:16.632Z
license_url
https://creativecommons.org/licenses/by/4.0/
methodology
Rows aggregate the benchmark-runs dataset exactly as the leaderboard pages do, over each benchmark's non-zero-score runs from the newest 90 days: accuracy is the task-weighted mean of per-run accuracy, accuracy_std_dev is the population standard deviation of per-run accuracy (null with a single run), and the per-task averages divide summed cost, generation time, and output tokens by summed tasks. Rows are ordered by accuracy descending, then model name.
metric_kind
tau2-bench Verified Airline · Accuracy
rank_method
competition rank among reported evaluations in identical conditions, including unlinked observations
sample_size
450
source_file
benchmark-leaderboard.json
source_page
https://openrouter.ai/benchmarks/tau2-bench-airline
measurements
accuracy
0.7481482222222222
accuracy_std_dev
0.020068464436510978
cost_per_task_usd
0.19089935333333333
output_tokens_per_task
11572.273333333333
latency_seconds_per_task
286.22822222222226
published_at
2026-11-16T01:46:02.653Z
snapshot_date
2026-11-16
source_revision
7ccaf17664252a3053dcb7df8ad0a46e72954a3b4524ddc1c8a7ef8a41b15568
transformations
TensorTrace preserves source results, converts latency to seconds and derives ranks within matching conditions. Individual runs have no derived rank.
metric_population
128
benchmark_evaluation
true
openrouter_permaslug
qwen/qwen3.8-2.4t-a95b-20260812
source model id
qwen/qwen3.8-2.4t-a95b-20260812
Original result ↗

OpenRouter Benchmarks

87.441%

source id
openRouterBenchmarks
leaderboard name
GPQA Diamond · Accuracy · auto-routing
dimension
knowledge
metric
GPQA Diamond · Accuracy
rank
41
sample size
1980
aggregation scope
status
published
harness
url
https://github.com/OpenRouterTeam/benchmark-harness
name
OpenRouter benchmark harness
direction
higherIsBetter
run_count
10
conditions
window
newest 90 days; successful visible nonzero runs
dataset
benchmark-leaderboard
benchmark
id
gpqa_diamond
name
GPQA Diamond
record_kind
aggregate
provider_name
auto-routing
attribution
Source: OpenRouter (openrouter.ai), licensed under CC BY 4.0: reuse and republish with attribution to OpenRouter.
last_run_at
2026-10-31T08:07:27.863Z
license_url
https://creativecommons.org/licenses/by/4.0/
methodology
Rows aggregate the benchmark-runs dataset exactly as the leaderboard pages do, over each benchmark's non-zero-score runs from the newest 90 days: accuracy is the task-weighted mean of per-run accuracy, accuracy_std_dev is the population standard deviation of per-run accuracy (null with a single run), and the per-task averages divide summed cost, generation time, and output tokens by summed tasks. Rows are ordered by accuracy descending, then model name.
metric_kind
GPQA Diamond · Accuracy
rank_method
competition rank among reported evaluations in identical conditions, including unlinked observations
sample_size
1980
source_file
benchmark-leaderboard.json
source_page
https://openrouter.ai/benchmarks/gpqa-diamond
measurements
accuracy
0.8744105999999999
accuracy_std_dev
0.014149500063253112
cost_per_task_usd
0.09050255303030302
output_tokens_per_task
14775.43686868687
latency_seconds_per_task
196.31064444444445
published_at
2026-11-16T01:46:02.653Z
snapshot_date
2026-11-16
source_revision
7ccaf17664252a3053dcb7df8ad0a46e72954a3b4524ddc1c8a7ef8a41b15568
transformations
TensorTrace preserves source results, converts latency to seconds and derives ranks within matching conditions. Individual runs have no derived rank.
metric_population
132
benchmark_evaluation
true
openrouter_permaslug
qwen/qwen3.8-2.4t-a95b-20260812
source model id
qwen/qwen3.8-2.4t-a95b-20260812
Original result ↗

Published prices

1–20 / 25

qwen/qwen3.8-2.4t-a95b ↗

$0.25

Inference provider
openRouter
Platform
OpenRouter
Unit
cache_read_tokens_per_million
Currency
USD
Billing conditions

qwen/qwen3.8-2.4t-a95b ↗

$2.00

Inference provider
openRouter
Platform
OpenRouter
Unit
input_tokens_per_million
Currency
USD
Billing conditions

qwen/qwen3.8-2.4t-a95b ↗

$6.00

Inference provider
openRouter
Platform
OpenRouter
Unit
output_tokens_per_million
Currency
USD
Billing conditions

Qwen/Qwen3.8-2.4T-A95B ↗

$0.25

Inference provider
siliconFlow
Platform
SiliconFlow
Unit
cache_read_tokens_per_million
Currency
USD
Billing conditions
Billing label
Cached input tokens (per 1M)

Qwen/Qwen3.8-2.4T-A95B ↗

$2.00

Inference provider
siliconFlow
Platform
SiliconFlow
Unit
input_tokens_per_million
Currency
USD
Billing conditions
Billing label
Input tokens (per 1M)

Qwen/Qwen3.8-2.4T-A95B ↗

$6.00

Inference provider
siliconFlow
Platform
SiliconFlow
Unit
output_tokens_per_million
Currency
USD
Billing conditions
Billing label
Output tokens (per 1M)

alibaba/qwen3.8-2.4t-a95b ↗

$0.20

Inference provider
vercelAIGatewayCatalog
Platform
deepinfra
Unit
cache_read_tokens_per_million
Currency
USD
Billing conditions
endpoint
deepinfra | alibaba/qwen3.8-2.4t-a95b
model_type
language
quantization
fp4
billing_scope
via Vercel AI Gateway
Billing label
Cached input / 1M
Pricing via
Vercel AI Gateway

alibaba/qwen3.8-2.4t-a95b ↗

$2.00

Inference provider
vercelAIGatewayCatalog
Platform
deepinfra
Unit
input_tokens_per_million
Currency
USD
Billing conditions
endpoint
deepinfra | alibaba/qwen3.8-2.4t-a95b
model_type
language
quantization
fp4
billing_scope
via Vercel AI Gateway
Billing label
Input tokens / 1M
Pricing via
Vercel AI Gateway

alibaba/qwen3.8-2.4t-a95b ↗

$6.00

Inference provider
vercelAIGatewayCatalog
Platform
deepinfra
Unit
output_tokens_per_million
Currency
USD
Billing conditions
endpoint
deepinfra | alibaba/qwen3.8-2.4t-a95b
model_type
language
quantization
fp4
billing_scope
via Vercel AI Gateway
Billing label
Output tokens / 1M
Pricing via
Vercel AI Gateway

alibaba/qwen3.8-2.4t-a95b ↗

$0.25

Inference provider
vercelAIGatewayCatalog
Platform
fireworks
Unit
cache_read_tokens_per_million
Currency
USD
Billing conditions
endpoint
fireworks | alibaba/qwen3.8-2.4t-a95b
model_type
language
billing_scope
via Vercel AI Gateway
Billing label
Cached input / 1M
Pricing via
Vercel AI Gateway

alibaba/qwen3.8-2.4t-a95b ↗

$2.00

Inference provider
vercelAIGatewayCatalog
Platform
fireworks
Unit
input_tokens_per_million
Currency
USD
Billing conditions
endpoint
fireworks | alibaba/qwen3.8-2.4t-a95b
model_type
language
billing_scope
via Vercel AI Gateway
Billing label
Input tokens / 1M
Pricing via
Vercel AI Gateway

alibaba/qwen3.8-2.4t-a95b ↗

$6.00

Inference provider
vercelAIGatewayCatalog
Platform
fireworks
Unit
output_tokens_per_million
Currency
USD
Billing conditions
endpoint
fireworks | alibaba/qwen3.8-2.4t-a95b
model_type
language
billing_scope
via Vercel AI Gateway
Billing label
Output tokens / 1M
Pricing via
Vercel AI Gateway

alibaba/qwen3.8-2.4t-a95b ↗

$0.25

Inference provider
vercelAIGatewayCatalog
Platform
gmicloud
Unit
cache_read_tokens_per_million
Currency
USD
Billing conditions
endpoint
gmicloud | alibaba/qwen3.8-2.4t-a95b
model_type
language
quantization
fp8
billing_scope
via Vercel AI Gateway
Billing label
Cached input / 1M
Pricing via
Vercel AI Gateway

alibaba/qwen3.8-2.4t-a95b ↗

$2.50

Inference provider
vercelAIGatewayCatalog
Platform
gmicloud
Unit
cache_write_tokens_per_million
Currency
USD
Billing conditions
endpoint
gmicloud | alibaba/qwen3.8-2.4t-a95b
model_type
language
quantization
fp8
billing_scope
via Vercel AI Gateway
Billing label
Cache write / 1M
Pricing via
Vercel AI Gateway

alibaba/qwen3.8-2.4t-a95b ↗

$2.00

Inference provider
vercelAIGatewayCatalog
Platform
gmicloud
Unit
input_tokens_per_million
Currency
USD
Billing conditions
endpoint
gmicloud | alibaba/qwen3.8-2.4t-a95b
model_type
language
quantization
fp8
billing_scope
via Vercel AI Gateway
Billing label
Input tokens / 1M
Pricing via
Vercel AI Gateway

alibaba/qwen3.8-2.4t-a95b ↗

$6.00

Inference provider
vercelAIGatewayCatalog
Platform
gmicloud
Unit
output_tokens_per_million
Currency
USD
Billing conditions
endpoint
gmicloud | alibaba/qwen3.8-2.4t-a95b
model_type
language
quantization
fp8
billing_scope
via Vercel AI Gateway
Billing label
Output tokens / 1M
Pricing via
Vercel AI Gateway

alibaba/qwen3.8-2.4t-a95b ↗

$0.25

Inference provider
vercelAIGatewayCatalog
Platform
novita
Unit
cache_read_tokens_per_million
Currency
USD
Billing conditions
endpoint
novita | alibaba/qwen3.8-2.4t-a95b
model_type
language
billing_scope
via Vercel AI Gateway
Billing label
Cached input / 1M
Pricing via
Vercel AI Gateway

alibaba/qwen3.8-2.4t-a95b ↗

$2.00

Inference provider
vercelAIGatewayCatalog
Platform
novita
Unit
input_tokens_per_million
Currency
USD
Billing conditions
endpoint
novita | alibaba/qwen3.8-2.4t-a95b
model_type
language
billing_scope
via Vercel AI Gateway
Billing label
Input tokens / 1M
Pricing via
Vercel AI Gateway

alibaba/qwen3.8-2.4t-a95b ↗

$6.00

Inference provider
vercelAIGatewayCatalog
Platform
novita
Unit
output_tokens_per_million
Currency
USD
Billing conditions
endpoint
novita | alibaba/qwen3.8-2.4t-a95b
model_type
language
billing_scope
via Vercel AI Gateway
Billing label
Output tokens / 1M
Pricing via
Vercel AI Gateway

alibaba/qwen3.8-2.4t-a95b ↗

$0.25

Inference provider
vercelAIGatewayCatalog
Platform
togetherai
Unit
cache_read_tokens_per_million
Currency
USD
Billing conditions
endpoint
togetherai | alibaba/qwen3.8-2.4t-a95b
model_type
language
quantization
fp4
billing_scope
via Vercel AI Gateway
Billing label
Cached input / 1M
Pricing via
Vercel AI Gateway

Provider specifications & access

1–2 / 2

Qwen3.8 2.4T A95B ↗

source model id
alibaba/qwen3.8-2.4t-a95b
hosting
vercel-gateway
platforms
  • Vercel AI Gateway
context length
262144
max output tokens
128000
architecture
input_modalities
  • text
  • image
output_modalities
  • text
capabilities
  • promptCaching
  • reasoning
  • structuredOutputs
  • toolCalling
  • visionInput

Qwen3.8-2.4T-A95B ↗

source model id
Qwen/Qwen3.8-2.4T-A95B
hosting
siliconflow-hosted
platforms
  • SiliconFlow
max output tokens
131072
architecture
input_modalities
  • text
output_modalities
  • text
capabilities
  • promptCaching
  • reasoning
  • structuredOutputs
  • toolCalling

Safety assessments

1–1 / 1

Foundation Model Transparency Index 2025 ↗

26 / 100

source id
stanfordFMTI
canonical model id
qwen/qwen3.8-2.4t-a95b-20260812
taxonomy
transparency
hazard category
transparency
hazard name
Organization transparency
policy area
Alibaba
assessment type
organization-level disclosure
metric
disclosure_score
result label
26 / 100
sample size
100
updated at
2026-11-09T18:08:05.227Z

Source specification observations

1–6 / 6

openRouter ↗

field
context_length
value
1048576
normalized value
1048576
scope
source-model
source id
openRouter
confidence
1

siliconFlow ↗

field
context_length
value
1049K
scope
source-model
source id
siliconFlow
confidence
1

vercelAIGatewayCatalog ↗

field
context_length
value
262144
normalized value
262144
scope
source-model
source id
vercelAIGatewayCatalog
confidence
1

openRouter ↗

field
max_output_tokens
value
131072
normalized value
131072
scope
source-model
source id
openRouter
confidence
1

siliconFlow ↗

field
max_output_tokens
value
131K
scope
source-model
source id
siliconFlow
confidence
1

vercelAIGatewayCatalog ↗

field
max_output_tokens
value
128000
normalized value
128000
scope
source-model
source id
vercelAIGatewayCatalog
confidence
1

Original sources

1–13 / 13

SiliconFlow · Model page ↗

Hugging Face · Model reference ↗

OpenRouter · Model page ↗

SiliconFlow · Playground ↗

Vercel AI Gateway · Model page ↗

SiliconFlow · API documentation ↗

Epoch AI · Catalog ↗

GitHub · Catalog ↗

Hugging Face · Catalog ↗

OpenRouter · Catalog ↗

OpenRouter · Catalog ↗

SiliconFlow · Catalog ↗

Vercel AI Gateway · Catalog ↗

Source links

1–9 / 9

Explore more models

Models

Loading
EVIDENCE / COMPARISON

Benchmarks

Explore every benchmark

Loading

These charts contain model-linked results. The charts above also include unlinked source profiles. Compare scores within each evaluation. Model quality, usage, transparency and system performance measure different things.

Install TensorTrace