Dev Tools
litgpt
20+ high-performance LLMs implemented from scratch, with proven recipes to pretrain, finetune and deploy.
★ 13.7k starsApache-2.0
View guide →A high-performance AI pipeline engine - build LLM workflows as portable .pipe files
rocketride-server is an open-source dev tools project by rocketride-org (rocketride-org/rocketride-server on GitHub). It has 17,818 GitHub stars and 7,606 forks, is released under the MIT license, and was last updated on GitHub on 8 October 2026.
Decide the flow - source, an LLM node, a response - for one concrete AI feature.
Install the Python or TypeScript SDK and the VS Code extension for visual building and debugging.
Drop an llm node into the pipeline and add one provider API key.
Execute the pipeline locally, inspect it with the built-in observability, then deploy the same file anywhere.
A high-performance core executes pipelines built from 50+ extensible nodes.
Design and debug workflows from your IDE with the included VS Code extension.
Connects 13+ model providers and 8+ vector databases as pipeline nodes.
Pipelines can include agent nodes with tools alongside RAG and data-processing nodes.
The same .pipe file runs on Cloud, on-prem, or as a local process in your IDE with zero lock-in.
Inspect pipeline runs while developing and in production.
Get the source code from GitHub and enter the project directory.
git clone https://github.com/rocketride-org/rocketride-server.git
cd rocketride-serverAdd RocketRide to your Python project (a TypeScript SDK is also published on npm).
pip install rocketrideFollow the README's first-pipe walkthrough: source, an llm node with one API key, and a response.
# Follow 'Your first AI feature' in the README, then try examples/rag-pipeline.pipeIt is maintained by the RocketRide project (rocketride-org) on GitHub.
The on-prem engine is free and MIT licensed. RocketRide Cloud is a paid managed hosting option with the same portable pipelines.
Pipelines that call hosted models need your own provider API keys; local components such as a local Qdrant instance can run without them.
Compared with code-only frameworks, RocketRide is a visual, engine-based approach with portable pipeline files and IDE debugging; compared with hosted workflow tools, the same pipelines also run fully on-prem.
Run pip install rocketride, install the VS Code extension, and build the README's first pipe with one LLM node.
20+ high-performance LLMs implemented from scratch, with proven recipes to pretrain, finetune and deploy.
An AI compute engine: a distributed runtime plus libraries for scaling ML and LLM workloads
A high-performance coding agent harness in Rust - fast TUI, swarms and every provider
Turn domain documents into high-quality LLM fine-tuning, RAG and evaluation datasets through a visual UI.
A portable home for your AI agent — memory, keys and automations on Cloudflare.
A GPU-accelerated, differentiable physics engine on NVIDIA Warp for robotics simulation.
The most-starred other Dev Tools projects in the LibreAI catalogue — the realistic shortlist if rocketride-server is not the right fit.
The open source coding agent
The model-definition framework behind modern AI — thousands of pretrained transformers for text, vision, and audio.
Anthropic's agentic coding tool that lives in your terminal and works across your whole codebase
Cross-platform desktop all-in-one manager for Claude Code, Codex, OpenCode and other coding agents
OpenAI's lightweight coding agent that runs locally in your terminal
Free and open — grab the source, star the repo, and start creating.