Unverified50% confidenceBenchmarkExact time
The anti-fraud model that Coinbase fine-tuned from open-source Qwen3.5-9B beat Claude Opus 4.5 on its own risk-control benchmark.
1
Sources
50%
Confidence
Short-term (~14 days)
Relevance
10/14/2026
First Seen
Valid until: 10/28/2026
Sources
Related Entities
Related Claims
UnverifiedQwen3.5, Qwen3.6, and Qwen3.8 gained faster prefill and speculative decoding support on MLX.61% similarUnverifiedMainstream models such as the Qwen3 series are already within the supported scope of the MLX runtime.61% similarUnverifiedTev1-4B-experimental is built on top of the Qwen3.5 4B base model.60% similarUnverifiedQwen3-0.8B is a small-parameter model that can be trained on a single consumer-grade GPU.59% similarPartially VerifiedQwen3.8-Flash-Next adopts a Mixture of Experts (MoE) architecture with 125B total parameters and 6B activated parameters.59% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/1027821API
curl https://kongchang.com/api/v1/knowledge/claims/1027821MCP
get_claim(id=1027821)