Unverified50% confidenceBenchmarkExact time
On the public test set, Falcon-ASR's WER was 20.92%, Jais-ASR's was 22.85%, and Audar-ASR-V1-Turbo's was 23.17%.
1
Sources
50%
Confidence
Medium-term (~90 days)
Relevance
11/17/2026
First Seen
Valid until: 2/15/2027
Sources
Related Entities
Related Claims
UnverifiedFalcon-ASR achieved an average Word Error Rate (WER) of 20.92% across 6 Arabic test sets.68% similarUnverifiedOn the 120-hour test set, Sada-ASR achieved an overall word error rate (WER) of 11.69%, while the strongest comparison system, Microsoft MAI-Transcribe-2, achieved 13.51%.65% similarUnverifiedWhen using the Prompt-SLAM-ASR-7B model with a 2,000-word bias list, B-WER on the test-clean set was reduced by 23.4% relative to the CTC-Filter baseline.65% similarUnverifiedArtificial Analysis gave GPT-6.1 Sol a score of 52, only one point below Astra, with benchmark run costs at one-quarter of Astra's and 31% cheaper than 6 Sol.63% similarUnverifiedIn the official Oswald test, Astra scored 72.6%, while the previous generation Cell scored 65.7%.61% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/1178052API
curl https://kongchang.com/api/v1/knowledge/claims/1178052MCP
get_claim(id=1178052)