A curated list of papers related to constrained decoding of LLM, along with their relevant code and resources.
369
71 commits
updated Jul 27, 2026
README
Awesome-LLM-Constrained-Decoding
Towards reliable, controllable and more efficient generation with Large Language Models (LLMs)
A curated list of papers related to constrained decoding of LLM, along with their relevant code and resources.
Table of Contents
Libraries
| Library | Feature | Stars |
|---|---|---|
| guidance-ai/guidance | CFG, Regex, JSON Schema, Token Forcing, compatible with Transformers, LLAMA-CPP |  |
| outlines-dev/outlines | CFG, Unicode support, Hugging Face ecosystem, VLLM support |  |
| jxnl/instructor | Try-Reject-Repeat approach to ensure constraints are met |  |
| eth-sri/lmql | Regex support, various constraints, more powerful control flow |  |
| microsoft/aici | A general framework of LLM controller with native support for CFG, Regex, JSON Schema |  |
| noamgat/lm-format-enforcer | Regex, JSON Schema, Beam Search etc. |  |
| mlc-ai/xgrammar | CFG, careful system optimizations |  |
| uiuc-focal-lab/syncode | CFG generation that supports builtin grammars like JSON, Python, Go, and more |  |
| Dan-wanna-M/formatron | Regex, JSON Schema, CFG, etc |  |
| genlm/genlm-control | Arbitrary programmable syntactic and semantic constraints, Constrained decoding as posterior inference, Sequential Monte Carlo |  |
| epfl-dlab/transformers-CFG | CFG (EBNF Interface), Compatible with Transformers, Easy to extend for research |  |
| eth-sri/type-constrained-code-generation | TypeScript, including type-safety. |  |
| litelines | Regex, JSON Schema, Tool-calling, triggered structured generation, visualization of the finite state machine |  |
| eth-sri/constrained-diffusion | Diffusion LLMs, Multi-Region Infilling (CFG) |  |
| snowkylin/circuit-transformer | Generating logical circuits that are strictly equivalent to given Boolean functions. |  |
| structuredllm/itergen | CFG generation and backtracking to handle semantic constraints |  |
| epfl-dlab/jsonschemabench | A benchmarking framework for evaluating constrained decoding engines on JSON Schema. Supports Guidance, Outlines, XGrammar, OpenAI and more. |  |
| grammarllm | LLprefix, DPDA |  |
Disclaimer:
- The libraries listed above are not exhaustive and are subject to change.
- The features mentioned are 100% not exhaustive and I strongly recommend checking the respective repositories for more details.
- The libraries are listed by the Github stars
- If you are the author of a library and would like to add or update the information, please open an issue or submit a pull request.
Papers
Papers with are newly added papers (not necessarily newly published papers).
Constrained Decoding for Autoregressive Models
Constrained Decoding for Diffusion Models
| Date | Paper | Publication |
|---|---|---|
| 2025-08 | Constrained Decoding of Diffusion LLMs with Context-Free Grammars | Preprint |
| 2025-05 | DINGO: Constrained Inference for Diffusion LLMs | Preprint |
| 2025-03 | Constrained Language Generation with Discrete Diffusion Models | Preprint |
Benchmark & Datasets & Evaluation
| Date | Paper | Publication |
|---|---|---|
| 2026-01 | StructEval: Benchmarking LLMs' Capabilities to Generate Structural Outputs | TMLR |
| 2025-08 | JSON-mode Eval Cleaned/Extended, SMILES Eval and HumanEval MRI C++ | HF Hub |
| 2025-01 | JsonSchemaBench: Generating Structured Outputs from Language Models: Benchmark and Studies | ICML ES-FoMo Workshop |
| 2024-05 | COLLIE: Systematic Construction of Constrained Text Generation Tasks | ICLR |
| 2024-02 | JSON-mode Eval dataset | HF hub |
| 2023-12 | BenchCLAMP: A Benchmark for Evaluating Language Models on Syntactic and Semantic Parsing | NeurIPS Track on Datasets and Benchmarks |
| 2023-10 | Evaluating Large Language Models on Controlled Generation Tasks | Preprint |
| 2023-09 | Struc-Bench: Are Large Language Models Really Good at Generating Complex Structured Data? | Preprint |
| 2021-10 | NLV corpus | CHI |
| 2020-12 | CommonGen: A Constrained Text Generation Challenge for Generative Commonsense Reasoning | EMNLP Findings |
| 2018-09 | Spider: A large-scale human-labeled dataset for complex and cross-domain semantic parsing and text-to-SQL task | EMNLP |
Survey
| Date | Paper | Publication |
|---|---|---|
| 2024-04 | "We Need Structured Output": Towards User-centered Constraints on Large Language Model Output | Preprint |
Blog Posts
- The good, the bad, and the ugly of Gemini’s structured outputs by Dylan Castillo
- Leveraging Constrained Sampling for Fill-in-the-Middle Code Completion by nielstron
- Proper Well-Formedness for Finite LLM Sampling by nielstron
- LLM Decoding with Regex Constraints by Vivien
- Constrained Decoding is Posterior Inference by Saibo-creator
- Making Structured Generation Faster Than Unstructured
- Coding For Structured Generation with LLMs
- Beating GPT-4 with Open Source
- Prompt Efficiency - Using Structured Generation to get 8-shot performance from 1-shot.
- How fast can grammar-structured generation be?
- Structured Generation Improves LLM performance: GSM8K Benchmark
- Coalescence: making LLM inference 5x faster
- Constrained Decoding with Arbitrary Constraints is NP-hard
- LLMs are bad at returning code in JSON
- Say what I mean
- LLGuidance: Making Structured Outputs Go Brrr
Many of the blogs are written by Outlines team, many thanks to them for their great work! ❤️
Related Awesome Lists
Disclaimer
This list is not exhaustive and will be updated regularly. If you have any suggestions or want to add a paper, please feel free to open an issue or submit a pull request. We hope to include all relevant papers in this list.
Contributing
Contributions are welcome! Feel free to submit a pull request or open an issue. Please make sure to read the Contributing Guidelines before contributing.
Contributors
Similar repos
A curated list of papers related to constrained decoding of LLM, along with their relevant code and resources.
369
71 commits
updated Jul 27, 2026
README
Awesome-LLM-Constrained-Decoding
Towards reliable, controllable and more efficient generation with Large Language Models (LLMs)
A curated list of papers related to constrained decoding of LLM, along with their relevant code and resources.
Table of Contents
Libraries
| Library | Feature | Stars |
|---|---|---|
| guidance-ai/guidance | CFG, Regex, JSON Schema, Token Forcing, compatible with Transformers, LLAMA-CPP | |
| outlines-dev/outlines | CFG, Unicode support, Hugging Face ecosystem, VLLM support | |
| jxnl/instructor | Try-Reject-Repeat approach to ensure constraints are met | |
| eth-sri/lmql | Regex support, various constraints, more powerful control flow | |
| microsoft/aici | A general framework of LLM controller with native support for CFG, Regex, JSON Schema | |
| noamgat/lm-format-enforcer | Regex, JSON Schema, Beam Search etc. | |
| mlc-ai/xgrammar | CFG, careful system optimizations | |
| uiuc-focal-lab/syncode | CFG generation that supports builtin grammars like JSON, Python, Go, and more | |
| Dan-wanna-M/formatron | Regex, JSON Schema, CFG, etc | |
| genlm/genlm-control | Arbitrary programmable syntactic and semantic constraints, Constrained decoding as posterior inference, Sequential Monte Carlo | |
| epfl-dlab/transformers-CFG | CFG (EBNF Interface), Compatible with Transformers, Easy to extend for research | |
| eth-sri/type-constrained-code-generation | TypeScript, including type-safety. | |
| litelines | Regex, JSON Schema, Tool-calling, triggered structured generation, visualization of the finite state machine | |
| eth-sri/constrained-diffusion | Diffusion LLMs, Multi-Region Infilling (CFG) | |
| snowkylin/circuit-transformer | Generating logical circuits that are strictly equivalent to given Boolean functions. | |
| structuredllm/itergen | CFG generation and backtracking to handle semantic constraints | |
| epfl-dlab/jsonschemabench | A benchmarking framework for evaluating constrained decoding engines on JSON Schema. Supports Guidance, Outlines, XGrammar, OpenAI and more. | |
| grammarllm | LLprefix, DPDA | |
Disclaimer:
- The libraries listed above are not exhaustive and are subject to change.
- The features mentioned are 100% not exhaustive and I strongly recommend checking the respective repositories for more details.
- The libraries are listed by the Github stars
- If you are the author of a library and would like to add or update the information, please open an issue or submit a pull request.
Papers
Papers with are newly added papers (not necessarily newly published papers).
Constrained Decoding for Autoregressive Models
Constrained Decoding for Diffusion Models
| Date | Paper | Publication |
|---|---|---|
| 2025-08 | Constrained Decoding of Diffusion LLMs with Context-Free Grammars | Preprint |
| 2025-05 | DINGO: Constrained Inference for Diffusion LLMs | Preprint |
| 2025-03 | Constrained Language Generation with Discrete Diffusion Models | Preprint |
Benchmark & Datasets & Evaluation
| Date | Paper | Publication |
|---|---|---|
| 2026-01 | StructEval: Benchmarking LLMs' Capabilities to Generate Structural Outputs | TMLR |
| 2025-08 | JSON-mode Eval Cleaned/Extended, SMILES Eval and HumanEval MRI C++ | HF Hub |
| 2025-01 | JsonSchemaBench: Generating Structured Outputs from Language Models: Benchmark and Studies | ICML ES-FoMo Workshop |
| 2024-05 | COLLIE: Systematic Construction of Constrained Text Generation Tasks | ICLR |
| 2024-02 | JSON-mode Eval dataset | HF hub |
| 2023-12 | BenchCLAMP: A Benchmark for Evaluating Language Models on Syntactic and Semantic Parsing | NeurIPS Track on Datasets and Benchmarks |
| 2023-10 | Evaluating Large Language Models on Controlled Generation Tasks | Preprint |
| 2023-09 | Struc-Bench: Are Large Language Models Really Good at Generating Complex Structured Data? | Preprint |
| 2021-10 | NLV corpus | CHI |
| 2020-12 | CommonGen: A Constrained Text Generation Challenge for Generative Commonsense Reasoning | EMNLP Findings |
| 2018-09 | Spider: A large-scale human-labeled dataset for complex and cross-domain semantic parsing and text-to-SQL task | EMNLP |
Survey
| Date | Paper | Publication |
|---|---|---|
| 2024-04 | "We Need Structured Output": Towards User-centered Constraints on Large Language Model Output | Preprint |
Blog Posts
- The good, the bad, and the ugly of Gemini’s structured outputs by Dylan Castillo
- Leveraging Constrained Sampling for Fill-in-the-Middle Code Completion by nielstron
- Proper Well-Formedness for Finite LLM Sampling by nielstron
- LLM Decoding with Regex Constraints by Vivien
- Constrained Decoding is Posterior Inference by Saibo-creator
- Making Structured Generation Faster Than Unstructured
- Coding For Structured Generation with LLMs
- Beating GPT-4 with Open Source
- Prompt Efficiency - Using Structured Generation to get 8-shot performance from 1-shot.
- How fast can grammar-structured generation be?
- Structured Generation Improves LLM performance: GSM8K Benchmark
- Coalescence: making LLM inference 5x faster
- Constrained Decoding with Arbitrary Constraints is NP-hard
- LLMs are bad at returning code in JSON
- Say what I mean
- LLGuidance: Making Structured Outputs Go Brrr
Many of the blogs are written by Outlines team, many thanks to them for their great work! ❤️
Related Awesome Lists
Disclaimer
This list is not exhaustive and will be updated regularly. If you have any suggestions or want to add a paper, please feel free to open an issue or submit a pull request. We hope to include all relevant papers in this list.
Contributing
Contributions are welcome! Feel free to submit a pull request or open an issue. Please make sure to read the Contributing Guidelines before contributing.