Search Results for "native" - Page 2

...The model supports full-stack development across web, mobile, and desktop platforms, covering the entire lifecycle from system design to testing and code review. With native serving speeds of up to 100 tokens per second, it completes complex agentic tasks significantly faster than previous versions while maintaining high token efficiency. M2.5 is built to be highly cost-effective, enabling continuous deployment of powerful AI agents at a fraction of the cost of other frontier models.

DevProjex turns any folder or Git repository into clean, token-efficient context for AI assistants and coding agents, and shows exactly what leaves your machine before you send it. Use it as a desktop app (GUI), a keyboard-first terminal workspace (TUI), a scriptable CLI for CI, or a read-only MCP server for Claude Code, Cursor, and other agents. An open-source alternative to Repomix, gitingest, and code2prompt, with a visual file tree and live preview. Smart Ignore drops build output,...

...Examples are stored as individual Markdown files, making the repository easy to browse and extend. Contributors can add their own favorite one-liners through pull requests. The project favors concise native JavaScript patterns rather than a packaged runtime library.

Aide is an AI software engineering assistant that provides conversational code generation, analysis, and refactoring directly within your IDE or through a local web interface. Built to act as a full-stack collaborator, it understands multi-file projects, detects dependency relationships, and can generate consistent updates across files and frameworks. It supports multiple backends—including OpenAI, Anthropic, and open-source models—and can route requests based on task type or latency...

...The model was pretrained on more than 35 trillion tokens and trained entirely on a large-scale cluster of domestically developed AI accelerators, demonstrating stable frontier-scale training without rollback events. LongCat-2.0 introduces LongCat Sparse Attention and extensive 1M-context training, enabling native processing of million-token inputs for long-document analysis, repository-scale coding, and complex multi-step reasoning. Dedicated post-training further strengthens coding and agent performance, producing competitive benchmark results against leading proprietary models.

...Architecturally, it uses a 1T-parameter Mixture-of-Experts design with 32B activated parameters, 61 layers, 384 experts, a 256K-token context window, and a MoonViT vision encoder. The model supports image and video input, native INT4 quantization, interleaved thinking, and multi-step tool calling. It also forces preserve-thinking mode by default, retaining full reasoning context across multi-turn interactions to improve coding-agent consistency. K2.7 Code is recommended for use through Kimi Code CLI and can be deployed with vLLM, SGLang, or KTransformers.