Share your thoughts, 1 month free Claude Pro on us
See more
Home
/
Benchmarks
Stance Detection on Multilingual Fact-Checking 28 languages (test)
Loading...
3.7
Median Latency (ms)
mmBERT-base fine-tuned (Factiverse)
-42.272
268.039
578.35
888.661
Jun 7, 2026
Median Latency (ms)
P95 Latency (ms)
Throughput (items/s)
Slowdown Factor
Macro F1
EN Macro F1
Cost ($/1k claims)
Updated 1mo ago
Evaluation Results
Method
Method
Links
Median Latency (ms)
P95 Latency (ms)
Throughput (items/s)
Slowdown Factor
Macro F1
EN Macro F1
Cost ($/1k claims)
mmBERT-base fine-tuned (Factiverse)
Hosting=self-hosted, E...
2026.06
3.7
4.2
267.6
1
62.05
80.36
0.008
Qwen3-8b
Hosting=self-hosted
2026.06
135.8
156.6
7.2
37
60.16
61.56
0.0289
GPT-5.2 (API)
Hosting=API, Evaluatio...
2026.06
1,006
1,451
1
272
64.57
67.05
0.7
Claude Opus 4.6 (API)
Hosting=API, Evaluatio...
2026.06
1,153
2,199
0.7
312
61.22
60.52
5.4
Feedback
Search any
task
Search any
task