Guidance, maintained under the guidance-ai organization (the project originated at Microsoft), is a programming model for controlling generation. Instead of sending one prompt and parsing a blob back, you write a template that interleaves fixed text, control flow, and constrained model calls, where each generated span can be forced to match a regex, a grammar, or a JSON Schema. The result is output that is structured by construction. Where Outlines, XGrammar, and llguidance are the enforcement engines, Guidance is the developer-facing layer: it gives the authoring model and delegates the token masking to an engine (including llguidance). Instructor occupies a nearby niche but is organized around returning typed objects from a normal call; Guidance is organized around scripting the generation itself. It matters because it changed how people think about prompting open models: from coaxing free text into shape to programming the output structure directly, which is more reliable and is the pattern much of the structured-output tooling now follows.
The Stack · Runtime · Open source
Guidance
A programming model for constrained generation that interleaves control flow with model calls and enforces regex, grammars, and JSON Schema.
Sources
- Guidance on GitHub https://github.com/guidance-ai/guidance
Want a follow-up? Ask the chat about Guidance in context. It will compare to siblings at the same layer and ground every claim in the wiki.
Other projects at the Runtime layer
12 siblings · ordered open first
- vLLM Open source
Dominant open production inference engine; PagedAttention and continuous batching; NVIDIA / AMD / Intel / TPU support.
- SGLang Open source
RadixAttention plus structured generation; from the LMSYS team; gains for shared-prefix and agent workloads.
- llama.cpp Open source
Georgi Gerganov's local-first inference engine; defines the GGUF format; the on-device standard.
- Ollama Open source
Local model runner; Docker-style UX over llama.cpp; the easiest way to run open weights on your machine.
- Text Generation Inference (TGI) Open source
HuggingFace's production inference server; maintenance mode in 2026 as vLLM became the standard.
- MLC-LLM Open source
Cross-platform compilation (TVM-based); the 'LLM in your browser' or 'on your phone' standard.
- Outlines Open source
Structured-output library using finite-state-machine guided decoding to force model output to match a regex, JSON Schema, or grammar.
- XGrammar Open source
Fast grammar-based constrained-decoding engine, used as a structured-output backend by several open serving engines.
- llguidance Open source
Low-level Rust constrained-decoding engine that enforces grammars and JSON Schema at high throughput; powers Guidance.
- Instructor Open source
Library for getting Pydantic-typed structured output from LLMs across many providers, built on function calling and JSON mode with automatic retries.
- LM Format Enforcer Open source
Token-filtering library that enforces a JSON Schema or regex during generation; integrates with Transformers, vLLM, and llama.cpp.
- TensorRT-LLM Source available
NVIDIA's closed-runtime counterpart; fastest on NVIDIA hardware; depends on closed CUDA kernels and the proprietary TensorRT compiler.