Unverified50% confidenceReleaseExact time
ModelScope announced the open-sourcing of Spark X 2.64B, a 4-billion-parameter model supporting a 1-million-token context and offering three modes: no-thinking, low-budget, and high-budget.
1
Sources
50%
Confidence
Medium-term (~90 days)
Relevance
10/21/2026
First Seen
Valid until: 1/19/2027
Sources
Related Entities
Related Claims
UnverifiediFlytek open-sourced two small models of Spark X 2.5, the 4B and 1.7B, targeting edge-side agent scenarios, employing a hybrid attention architecture with native support for up to 1 million token context.64% similarUnverifiedreef's OpenClaw-RL recipe uses 72 GSM8K problems as the task stream, with Qwen3-4B serving simultaneously as the policy model and the process reward model, Qwen3-32B as the student model, using seven GPUs, and setting a threshold that must be reached before an update is accepted.62% similarUnverifiedXingchen 4.0 29B is compatible with fine-tuning tools such as MindFormers and LLaMA-Factory, as well as inference frameworks such as SGLang and vLLM.61% similarUnverifiedChina Telecom's open-sourced Xingchen series models adopt a MoE architecture, support 256K context extendable to 512K, and are among the first large models trained entirely on Ascend 910C.60% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/1081844API
curl https://kongchang.com/api/v1/knowledge/claims/1081844MCP
get_claim(id=1081844)