Unverified50% confidenceReleaseExact time

The implementation, including expert scheduling and DSpark integration, has been open-sourced in llama.cpp-omni.

1

Sources

50%

Confidence

Medium-term (~90 days)

Relevance

10/21/2026

First Seen

Valid until: 1/19/2027

Sources

iPhone Runs 35B MoE at 20.5 tok/s: The Scheduling Magic That Skips No Experts

redditr/LocalLLaMA10/21/2026

Related Entities

llama.cpp-omniproduct OpenBMBcompany DSparkmodel

Related Claims

UnverifiedVeloce-Engine is compatible with any OpenAI-compatible API, Anthropic endpoints, and local models served via Ollama or llama.cpp.68% similar UnverifiedOpenRouter是2023年兴起的LLM统一路由服务，通过单一API端点聚合了OpenAI、Anthropic、Google、Meta等数十家模型提供商的能力64% similar

Cite This Claim

Stable URI

https://kongchang.com/claim/1081661

API

curl https://kongchang.com/api/v1/knowledge/claims/1081661

MCP

get_claim(id=1081661)