Unverified50% confidenceReleaseExact time
French AI lab KainTai has released the REKI vision-language-action model, with only 2.4 billion parameters, and all weights have been open-sourced.
1
Sources
50%
Confidence
Short-term (~14 days)
Relevance
10/18/2026
First Seen
Valid until: 11/1/2026
Sources
DeepSeek and StepFun Open-Source on the Same Day, While Websites Shut Out AI Agents
bilibiliSKC智能知识管理系统10/18/2026
Related Entities
Related Claims
VerifiedVLA models integrate visual perception, language understanding, and robot action planning into a unified model, typically using a pretrained large vision-language model as the backbone, with representative works including Google's RT-2 and OpenVLA.64% similarUnverifiedQwen2.5-VL 3B is a multimodal vision-language model released by Alibaba's Tongyi Qianwen team, with roughly 3 billion parameters, runnable on consumer laptops, and commonly used for OCR, chart understanding, and structured document extraction.64% similarUnverifiedThe computer vision model behind iNaturalist has been open-sourced.61% similarVerifiedOpen-source multimodal models LLaVA, InternVL, and Qwen-VL typically adopt a 'vision encoder + projection layer + language model' architecture.61% similarUnverifiedOpenAI launched Function Calling in 2023, after which Anthropic's Tool Use, Google DeepMind's ReAct framework, and others successively refined the standard paradigm for model interaction with external systems.61% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/1059852API
curl https://kongchang.com/api/v1/knowledge/claims/1059852MCP
get_claim(id=1059852)