I spent a couple years building a high performance, expressive library for structured outputs in LLMs. Our library is used by OpenAI for structured outputs on the hosted API. Happy to answer questions on how this works:
TL;DR instead of just getting a token and seeing if it would be accepted by the parser, you can actually zero-out probabilities for all invalid tokens, and do the computation for this in parallel at effectively zero cost:
> Here, compute_mask() can run on the CPU during the time it would be normally just waiting for the GPU to finish. The line prob[~mask] = 0.0 would normally be fused into the softmax kernel in the last stage of the LLM, with negligible overhead. Therefore, as long as the compute_mask() function completes faster than the LLM forward pass and parser.consume() is negligible (typically follows from compute_mask() speed), the constrained generation will be as fast as the unconstrained one.
I'm curious - have there been any research/conversations about pushing masking even earlier in the pipeline? In theory, there's a fair amount of compute that goes into computing the probability of tokens that will end up being masked away anyways.
Well, thank you for that; from a quick skim of Guidance, it looks like it is used when interfacing with the model directly - i.e. if I want to use Guidance I can't simply send input to my local Ollama instance, I have to stand up a small Python program that loads the model, accepts input from the user, push the user input tokens into the model, and for each output token, reject it if it fails some criteria.
Is this correct? If so, it means that the current way LLMs are interfaced with (via stdin/stout or an HTTP endpoint) can't be used with something like Guidance, correct?
"The constraint system offered by Guidance is extremely powerful. It can ensure that the output conforms to any context free grammar (so long as the backend LLM has full support for Guidance). More on this below." --from https://github.com/guidance-ai/guidance/
I didn't find any more on that comment below. Is there a list of supported LLMs?
We have support for Huggingface Transformers, llama.cpp, vLLM, SGLang, and TensorRT-LLM, along with some smaller providers (e.g. mistral.rs). Using any of these libraries as an inference host means you can use an OSS model with the guidance backend for full support. Most open source models will run on at least one of these backends (with vLLM probably being the most popular hosted solution, and transformers/llama.cpp being the most popular local model solutions)
We're also the backend used by OpenAI/Azure OpenAI for structured outputs on the closed source model side.
Guidance is genuinely impressive for anyone wrangling LLM output. The ability to map grammar constraints so efficiently at inference solves so many subtle issues—tokenization headaches being just one. Curious if you've benchmarked adoption for JSON vs. custom grammars among production teams? Anecdotally, JSON's become the baseline, but custom grammars unlock way more nuanced applications.
Great question re: adoption...it's definitely dominated by JSON. Most API providers have standardized on JSON outputs, so application teams have started building shims that map other formats to JSON and back. Similarly, with models heavily being post-trained to generate "good" JSON, I think there's a better model-constraint alignment story with JSON than most arbitrary grammars.
That said, internally, we experiment quite a lot with custom grammars all across the stack. It's more complicated to write a grammar than a JSON schema (though LMs are very good at grammar writing now) and more error prone to debug, but it can help significantly in certain cases (e.g. having models write custom DSLs not commonly found on the internet, at various parts of a model training pipeline, etc. etc.). I'm hoping that with the right tooling around it, the broader community will start nudging beyond JSON.
To that end, the python guidance library is really an attempt to make writing grammars more friendly to a python programmer. More to be done here of course!
We did quite a thorough benchmarking of various structured decoding providers in one of our papers: https://arxiv.org/abs/2501.10868v3 , measuring structured outputs providers on performance, constraint flexibility, downstream task accuracy, etc.
Happy to chat more about the benchmark. Note that these are a bit out of date though, I'm sure many of the providers we tested have made improvements (and some have switched to wholesale using llguidance as a backend)
I'm trying to write a really large book. I have a lot of material that I'm using RAG to help manage. I put into my prompts the top RAG cosine scores with some summaries of characters and previous chapters and scene sketches. I get scenes out and then work them over. LLMs are really helpful for my disability and have allowed me to make any progress at all on this.
Is your thing something I should look into for helping keep track of my material. I'm using Excel sheets and crappy python code right now.
Im pretty sure your stuff is some super technical backend thingy, but I figured I'd shoot my shot here. Thanks for any and all info, I appreciate it
I've been curious about grammar support for non-JSON applications. (i.e., I have some use cases where XML is more natural and easier to parse but Pydantic seems to assume you should only work with JSON.) Would guidance be able to handle this use case?
In general I find that matching the most natural format for a document outperforms waiting for the big model trainers to convince the model that the format you want is a valid structure, so anything that lets me interweave structured and unstructured generation is very interesting to me right now.
guidance can handle many context-free grammars. We use an Earley parser under the hood (https://en.wikipedia.org/wiki/Earley_parser) which gives us significant flexibility boosts over alternative approaches that use weaker parsers (and went through lots of effort to make Earley parsing fast enough to not slow down LM inference). However, XML is not perfectly context-free, though with some basic assumptions you can make it CF.
The annoying bit with grammars is that they are unfortunately a bit complex to write properly. Fortunately language models are getting better at this, so hopefully to get an XML grammar, you can get most of the way there with just a GPT-5 prompt. Suppose it would be a good idea to have a better pre-built set of popular grammars (like a modified XML) in guidance so that we cut this headache out for users...!
This is a great writeup! There was a period where reliable structured output was a significant differentiator and was the 'secret sauce' behind some companies success. A NL->SQL company I am familiar with comes to mind. Nice to see this both public and supported by a growing ecosystem of libraries.
One statement surprised me was that the author thinks "models over time will just be able to output JSON perfectly without the need for constraining over time."
I'm not sure how this conclusion was reached. "Perfectly" is a bar that probabilistic sampling cannot meet.
Thank you! Maybe not "perfect" but near-perfect is something we can expect. Models like the Osmosis structure which just structure data inspired some of that thinking (https://ollama.com/Osmosis/Osmosis-Structure-0.6B). Historically, JSON generation has been a latent capability of a model rather than a trained one, but that seems to be changing. gpt-oss was particularly trained for this type of behavior and so the token probabilities are heavily skewed to conform to JSON. Will be interesting to see the next batch of models!
You're spot on about the "perfect" JSON bar being unreachable for now. The only consistently reliable method I've seen in the wild is some form of constrained decoding or grammar enforcement—bit brittle, but practical. Sampling will always be fuzzy unless the architecture fundamentally shifts. Anyone claiming zero-validity issues is probably glossing over a ton of downstream QA work.
We’ve had a lot of success implementing schema-aligned parsing in BAML, a DSL that we’ve built to simplify this problem.
We actually don’t like constrained generation as approach - among other issues it limits your ability to use reasoning - and instead the technique we’re using is algorithm-driven error-tolerant output parsing.
I've a related observation. In my experience the amount of hallucinated urls with structured output (think of a field `url` or `link`) is pretty high. Especially compared to the alternative approach, where you let the llm generate text and then use a second llm to convert the text into the desired structured format.
With structured output, it's like the llm is forced to answer in a very specific way. So if there is no url for the given field, it makes up the url.
Here a related quote from the article:
> Structured outputs builds on top of sampling by constraining the model's output to a specific format.
What I've found is that it is very important to make structured outputs as easy for the LLM as possible. This means making your schemas LLM-friendly instead of programmer-friendly.
E.g. if the LLM hallucinates non-existing URLs, you may add a boolean "contains_url" field to your entity's JSON schema, placing it before the URL field itself. This way, the URL extraction is split into two simpler steps, checking if the URL is there and actually extracting it. If the URL is missing, the `"contains_url": false` field in the context will strongly urge the LLM to output an empty string there.
This also comes up with quantities a lot. Imagine you're trying to sort job adverts by salary ranges, which you extract via LLm. . These may be expressed as monthly instead of annual (common in some countries), in different currencies, pre / post tax etc.
Instead of having an `annual_pretax_salary_usd` field, which is what you actually want, but which the LLM is extremely ill-equipped to generate, have a detailed schema like `type: monthly|yearly, currency:str, low:float, high:float, tax: pre_tax|post_tax`.
That schema is much easier for an LLM to generate, and you can then convert it to a single number via straight code.
As you know, (most current) LLMs build text autoregressively. This allows them to generate text with _exactly_ the same distribution as the training data.
When you constrain LLM output at each token, that gives a completely different distribution from letting the LLM generate a full output and then doing something with that (trying again, returning an error, post-processing, etc).
E.g.: Suppose the LLM has a training set of (aa, ab, ab, ba), noting that "ab" appears twice. Suppose your valid grammar is the set (ab, ba). Then your output distributions are:
Baseline: {invalid: 25%, ab: 50%, ba: 25%}
Constrained: {invalid: 0%, ab: 75%, ba: 25%}
Note that _all_ the previously invalid outputs were dumped into the "ab" bucket, skewing the ratio between "ab" and "ba". That skew may or may not be desirable, but assuming the training process was any good it's likely undesirable.
You've observed it in URLs, but I see it in JSON output as well. LLMs like to truncate long strings from time to time, but when they do they're more likely to provide invalid JSON (adding an ellipsis at the end of the fragment and doing nothing else). If that truncation starts to happen in a constrained environment, a period is a valid character in a long string, and eventually the grammar constraint will force a closing quote to appear. The result is still garbage, but instead of a detectable parse failure you have an undetectable corrupt field.
> let the llm generate text and then use a second llm to convert the text into the desired structured format
this sounds similar to what they discussed in the article with regards to "thinking" models, i.e. let them generate their <think>blah blah</think> preamble first before starting to constrain the output to structured format
Google's Gemini API is a bit odd with structured outputs. If you specify an Application/JSON response mimetype, it will reliably respond with a consistent JSON output without any prompt engineering shenanigans. For my workflows, this setting plus providing a JSON Schema in the system prompt works even with complex schema.
The Gemini API has a canonical implementation of structured outputs where you can instead pass the JSON schema as a separate parameter to control the grammar more closely. However, this setting will reorder the JSON schema fields to be alphabetical beforehand, which is especially not desired behavior as the order of JSON fields in a schema is often very deliberate to control generation.
Have you tried techniques that don’t require modifying the LLM and the sampling strategy for structure outputs? For example, schema aligned passing, where you build error tolerance into the parser instead of coercing to a grammar.
It looks really slick, for us the reason we haven't adopted yet is it brings more tooling and configuration that overlaps with our existing system for prompt templates, schema definitions, etc. In the component where we couldn't rely on OpenAI structured outputs we experimented with TOML-formatted output, that ended up being reliable enough to solve the problem across many models without any new dependencies. I do think we'll revisit at some point as Boundary also provides incremental parsing of streaming outputs and may allow some cost optimization that is not easy right now.
If you can screen tokens against your grammar fast enough, you can build a bitmask over the entire token vocabulary and apply it right before sampling. As vocabulary sizes grow, this gets more complex to do in real time, but we (and other libraries) have found several optimizations to do this extremely quickly (eg for guidance, we detail some optimizations here https://github.com/guidance-ai/llguidance/blob/main/docs/opt...).
Other libraries work by essentially pre-computing all the masks for all possible generations, but of course you're restricted to working with simple grammars in this case (like a subset of regular expressions)
It's not expensive per-se; A single element-wise multiplication of the output vector.
The real "expense" is that you need to prepare masks for every element of your grammar as they are expensive to recompute as needed; LLM tokens do not cleanly map onto elements of your grammar. (Consider JSON: LLM tokens often combine various special characters such as curly braces, colons, and quotes.)
This isn't that hard to compute, it's just more work to implement.
Good question—some frameworks do apply the mask immediately, others defer for performance or implementation simplicity. Mask precomputation can get tricky with large vocabularies, especially if grammar elements span multiple tokens. Immediate masking is usually preferred, but optimizations kick in when you're juggling complicated grammars or working against throughput bottlenecks.
Hey! I'm the author of the post. We haven't optimized sampling yet so it's running linearly on the CPU. A lot of SOTA work either does this while the model is running the forward pass or does the masking on the GPU.
The greedy accept is so that the mask doesn't need to be computed. Planning to make this more efficient from either ends.
I was hoping to find some insights about why performance drops when using actual structured outputs. It's been a known problem. For example this paper "Let Me Speak Freely? A Study on the Impact of Format Restrictions on Performance of Large Language Models" says:
> Surprisingly, we observe a significant decline in LLMs’ reasoning abilities under format restrictions. Furthermore, we find that stricter format constraints generally lead to greater performance degradation in reasoning tasks.
That paper had some serious methodological issues and the results have been shown to be misunderstood/incorrect in the majority of cases. In fact, in many cases structured outputs have shown to improve the quality of the results from an LLM (at least in terms of evaluation performance). The team at behind the Outlines library released a response the covers the issues in details and provides more information about structured outputs [0].
Readit News
User friendly library that connects to lots of OSS model serving backends: https://github.com/guidance-ai/guidance/
Core Rust library written for high performance mask computation (written mostly by my collaborator @mmoskal): http://github.com/guidance-ai/llguidance
TL;DR instead of just getting a token and seeing if it would be accepted by the parser, you can actually zero-out probabilities for all invalid tokens, and do the computation for this in parallel at effectively zero cost:
> Here, compute_mask() can run on the CPU during the time it would be normally just waiting for the GPU to finish. The line prob[~mask] = 0.0 would normally be fused into the softmax kernel in the last stage of the LLM, with negligible overhead. Therefore, as long as the compute_mask() function completes faster than the LLM forward pass and parser.consume() is negligible (typically follows from compute_mask() speed), the constrained generation will be as fast as the unconstrained one.
I'm curious - have there been any research/conversations about pushing masking even earlier in the pipeline? In theory, there's a fair amount of compute that goes into computing the probability of tokens that will end up being masked away anyways.
Well, thank you for that; from a quick skim of Guidance, it looks like it is used when interfacing with the model directly - i.e. if I want to use Guidance I can't simply send input to my local Ollama instance, I have to stand up a small Python program that loads the model, accepts input from the user, push the user input tokens into the model, and for each output token, reject it if it fails some criteria.
Is this correct? If so, it means that the current way LLMs are interfaced with (via stdin/stout or an HTTP endpoint) can't be used with something like Guidance, correct?
Should work with any llama.cpp compatible model: https://github.com/sutt/innocuous
I didn't find any more on that comment below. Is there a list of supported LLMs?
We have support for Huggingface Transformers, llama.cpp, vLLM, SGLang, and TensorRT-LLM, along with some smaller providers (e.g. mistral.rs). Using any of these libraries as an inference host means you can use an OSS model with the guidance backend for full support. Most open source models will run on at least one of these backends (with vLLM probably being the most popular hosted solution, and transformers/llama.cpp being the most popular local model solutions)
We're also the backend used by OpenAI/Azure OpenAI for structured outputs on the closed source model side.
Great question re: adoption...it's definitely dominated by JSON. Most API providers have standardized on JSON outputs, so application teams have started building shims that map other formats to JSON and back. Similarly, with models heavily being post-trained to generate "good" JSON, I think there's a better model-constraint alignment story with JSON than most arbitrary grammars.
That said, internally, we experiment quite a lot with custom grammars all across the stack. It's more complicated to write a grammar than a JSON schema (though LMs are very good at grammar writing now) and more error prone to debug, but it can help significantly in certain cases (e.g. having models write custom DSLs not commonly found on the internet, at various parts of a model training pipeline, etc. etc.). I'm hoping that with the right tooling around it, the broader community will start nudging beyond JSON.
To that end, the python guidance library is really an attempt to make writing grammars more friendly to a python programmer. More to be done here of course!
I'm yet to see a thorough comparison of design, performance and reliability between these options (along with outlines etc)
Happy to chat more about the benchmark. Note that these are a bit out of date though, I'm sure many of the providers we tested have made improvements (and some have switched to wholesale using llguidance as a backend)
I'm trying to write a really large book. I have a lot of material that I'm using RAG to help manage. I put into my prompts the top RAG cosine scores with some summaries of characters and previous chapters and scene sketches. I get scenes out and then work them over. LLMs are really helpful for my disability and have allowed me to make any progress at all on this.
Is your thing something I should look into for helping keep track of my material. I'm using Excel sheets and crappy python code right now.
Im pretty sure your stuff is some super technical backend thingy, but I figured I'd shoot my shot here. Thanks for any and all info, I appreciate it
In general I find that matching the most natural format for a document outperforms waiting for the big model trainers to convince the model that the format you want is a valid structure, so anything that lets me interweave structured and unstructured generation is very interesting to me right now.
The annoying bit with grammars is that they are unfortunately a bit complex to write properly. Fortunately language models are getting better at this, so hopefully to get an XML grammar, you can get most of the way there with just a GPT-5 prompt. Suppose it would be a good idea to have a better pre-built set of popular grammars (like a modified XML) in guidance so that we cut this headache out for users...!
One statement surprised me was that the author thinks "models over time will just be able to output JSON perfectly without the need for constraining over time."
I'm not sure how this conclusion was reached. "Perfectly" is a bar that probabilistic sampling cannot meet.
We actually don’t like constrained generation as approach - among other issues it limits your ability to use reasoning - and instead the technique we’re using is algorithm-driven error-tolerant output parsing.
https://boundaryml.com/
I've a related observation. In my experience the amount of hallucinated urls with structured output (think of a field `url` or `link`) is pretty high. Especially compared to the alternative approach, where you let the llm generate text and then use a second llm to convert the text into the desired structured format.
With structured output, it's like the llm is forced to answer in a very specific way. So if there is no url for the given field, it makes up the url.
Here a related quote from the article:
> Structured outputs builds on top of sampling by constraining the model's output to a specific format.
E.g. if the LLM hallucinates non-existing URLs, you may add a boolean "contains_url" field to your entity's JSON schema, placing it before the URL field itself. This way, the URL extraction is split into two simpler steps, checking if the URL is there and actually extracting it. If the URL is missing, the `"contains_url": false` field in the context will strongly urge the LLM to output an empty string there.
This also comes up with quantities a lot. Imagine you're trying to sort job adverts by salary ranges, which you extract via LLm. . These may be expressed as monthly instead of annual (common in some countries), in different currencies, pre / post tax etc.
Instead of having an `annual_pretax_salary_usd` field, which is what you actually want, but which the LLM is extremely ill-equipped to generate, have a detailed schema like `type: monthly|yearly, currency:str, low:float, high:float, tax: pre_tax|post_tax`.
That schema is much easier for an LLM to generate, and you can then convert it to a single number via straight code.
As you know, (most current) LLMs build text autoregressively. This allows them to generate text with _exactly_ the same distribution as the training data.
When you constrain LLM output at each token, that gives a completely different distribution from letting the LLM generate a full output and then doing something with that (trying again, returning an error, post-processing, etc).
E.g.: Suppose the LLM has a training set of (aa, ab, ab, ba), noting that "ab" appears twice. Suppose your valid grammar is the set (ab, ba). Then your output distributions are:
Baseline: {invalid: 25%, ab: 50%, ba: 25%}
Constrained: {invalid: 0%, ab: 75%, ba: 25%}
Note that _all_ the previously invalid outputs were dumped into the "ab" bucket, skewing the ratio between "ab" and "ba". That skew may or may not be desirable, but assuming the training process was any good it's likely undesirable.
You've observed it in URLs, but I see it in JSON output as well. LLMs like to truncate long strings from time to time, but when they do they're more likely to provide invalid JSON (adding an ellipsis at the end of the fragment and doing nothing else). If that truncation starts to happen in a constrained environment, a period is a valid character in a long string, and eventually the grammar constraint will force a closing quote to appear. The result is still garbage, but instead of a detectable parse failure you have an undetectable corrupt field.
this sounds similar to what they discussed in the article with regards to "thinking" models, i.e. let them generate their <think>blah blah</think> preamble first before starting to constrain the output to structured format
The Gemini API has a canonical implementation of structured outputs where you can instead pass the JSON schema as a separate parameter to control the grammar more closely. However, this setting will reorder the JSON schema fields to be alphabetical beforehand, which is especially not desired behavior as the order of JSON fields in a schema is often very deliberate to control generation.
You can specify ordering in the Gemini API with propertyOrdering:
"propertyOrdering": ["recipeName", "ingredients"]
https://boundaryml.com/blog/schema-aligned-parsing
Why wouldn't we apply the mask immediately for the first sampling? Is this an optimization somehow, is masking expensive?
Other libraries work by essentially pre-computing all the masks for all possible generations, but of course you're restricted to working with simple grammars in this case (like a subset of regular expressions)
> is masking expensive?
It's not expensive per-se; A single element-wise multiplication of the output vector.
The real "expense" is that you need to prepare masks for every element of your grammar as they are expensive to recompute as needed; LLM tokens do not cleanly map onto elements of your grammar. (Consider JSON: LLM tokens often combine various special characters such as curly braces, colons, and quotes.)
This isn't that hard to compute, it's just more work to implement.
The greedy accept is so that the mask doesn't need to be computed. Planning to make this more efficient from either ends.
Human
4x1200 with 30 second rest
AI DSL output
Repeat 4 times:
- Run 1200 meters
- Rest 30 seconds
I hand wrote a recursive descent parser in Python to process DSL. Human speech to DSL is pretty effective with a simple prompt and some examples.
I created a tool that can program Garmin & Apple Watches for interval training based on what I wrote above.
https://speedystride.com
Looking for beta testers- please give it a try :)
> Surprisingly, we observe a significant decline in LLMs’ reasoning abilities under format restrictions. Furthermore, we find that stricter format constraints generally lead to greater performance degradation in reasoning tasks.
https://arxiv.org/abs/2408.02442v1
0. https://blog.dottxt.ai/say-what-you-mean.html