Moonshot AI releases PerceptionBench, a benchmark for evaluating atomic visual perception in multimodal large language models, with 3,000 verified questions across 10 atomic perceptual categories,…
Moonshot AI introduces Kimi K3, a 2.8T-parameter open model with native vision and a 1-million-token context window, available today on Kimi platforms and API, with weights to be released by July…
Moonshot AI is open-sourcing Kimi K2.6, a model with state-of-the-art coding, long-horizon execution, and agent swarm capabilities. It is available via Kimi.com, Kimi App, API, and Kimi Code.
Moonshot AI announced Agent Swarm, a new system where multiple AI agents can work in parallel, self-organize into teams with roles, and execute tasks like research or report writing faster than…
Moonshot AI releases WorldVQA, a benchmark of 3,500 image-question pairs to evaluate factual visual world knowledge in multimodal LLMs, with a focus on head vs. tail distribution.
Moonshot AI introduces Kimi K2.5, an open-source multimodal model with state-of-the-art coding and vision capabilities, and a self-directed agent swarm paradigm that can orchestrate up to 100…
Moonshot AI is open-sourcing the Kimi Vendor Verifier (KVV), a project that provides benchmarks and tools to verify the accuracy of inference implementations for open-source models, addressing issues…
Moonshot AI announces Kimi K2 Thinking, an open-source thinking model that reasons step by step while using tools, achieving state-of-the-art scores on HLE, BrowseComp, and SWE-Bench Verified…
Moonshot AI announces Kimi K2, a Mixture-of-Experts model with 32 billion activated parameters and 1 trillion total parameters, optimized for agentic tasks and open-sourced as Kimi-K2-Base and…
Moonshot AI released the MoonViT-SO-400M vision transformer model on Hugging Face under the MIT license for image feature extraction.