Top all-time · #3450
BAML
Company blog · boundaryml.com/blog · on HN as aaronvgprogrammingai
BAML is the company blog at boundaryml.com, writing about programming and ai. Since 2024, 21 different Hacker News readers have linked it 38 times, most often “BAML is Basically A Made-up Language”. No feed was found, so read it on the site itself.
Too few sampled comments (15) to say how often HN recommends it.?
Why HN links it
HN readers cite BAML's blog as a reference for schema-aligned parsing and structured LLM output. Comments mostly promote the BAML tool itself rather than judge the posts.
Summarised by Claude from a sample of 15 HN comments that link here. Each comment is marked recommending, critical or neutral (citing without judging); how this works.
BAML is a programming language for agents. Python and TypeScript were built for human productivity. BAML was designed with a different goal in mind.
Most-linked posts?
- BAML is Basically A Made-up Language6 people · 6 mentions · first linked 2025-09-23
- Prompts are using 4x more tokens than you need3 people · 3 mentions · first linked 2024-04-01
- /3 people · 3 mentions · first linked 2024-06-27
- Every Way to Get Structured Output from LLMs3 people · 3 mentions · first linked 2024-05-31
- Show HN: Beating OpenAI's structured outputs on cost, accuracy and speed2 people · 3 mentions · first linked 2024-08-14
- Show HN: Structured Outputs with Deepseek R12 people · 3 mentions · first linked 2025-01-21
- Structured outputs create false confidence2 people · 2 mentions · first linked 2025-12-15
What Hacker News says?
have you tried schema-aligned parsing yet? the idea is that instead of using JSON.parse, we create a custom Type.parse for each type you define. so if you want a: class Job { company: string[] } And the LLM happens to output: { "company": "Amazon" } We can upcast "Amazon" -> ["Amazon"] since you indicated that in your schema. https://www.boundaryml.com/blog/schema-aligned-parsing and since its only post processing,…
Comment linking /blog/sota-function-calling
on “Andrej Karpathy: Software in the era of AI [video]”Submitted Every Way to Get Structured Output from LLMs as a story · 169 points
HN discussionI have had a lot of success using BoundaryML ( https://www.boundaryml.com/ ) for this. They have also been super responsive for any of my questions.
Comment linking /
on “Structured Outputs in the API”Submitted Structured outputs create false confidence as a story · 136 points
HN discussionYou might also find Semantic Streaming interesting. It's t he same concept but applied to llm token streaming. It's used in BAML (the ai framework). https://www.boundaryml.com/blog/semantic-streaming I'm one of the developers of BAML.
Comment linking /blog/semantic-streaming
on “Progressive JSON”Submitted How to write a Zed extension for a made up language as a story · 8 points
HN discussionyea! even deepseek. Calling an external function / tool calling is really just a data extraction problem. say you have a tool: def calculator(expr: str) -> float then the model just needs to say: { "function": "calculator", "args": { "expr": "5 + 10" } } then in your code you can easily pass that to the "calculator" function and get the result, then hand the result back to the model. Making it feel like the model ca…
Comment linking /blog/deepseek-r1-function-calling
on “DeepScaleR: Surpassing O1-Preview with a 1.5B Model by Scaling RL”Submitted Show HN: Beating OpenAI's structured outputs on cost, accuracy and speed as a story · 8 points
HN discussionHey, this is Gabe from zenfetch. Been following you guys for a few months now since your first launch. I definitely resonate with all the problems you've described regarding celery shortcomings / other distributed task queues. We're on celery right now and have been through the ringer with various workflow platforms. Only reason we haven't switched to Hatchet is because we are finally in a stable place, though that…
Comment linking /
on “Launch HN: Hatchet (YC W24) – Open-source task queue, now with a cloud version”Submitted Show HN: Structured Outputs with Deepseek R1 as a story · 7 points
HN discussionHave 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. https://boundaryml.com/blog/schema-aligned-parsing
Comment linking /blog/schema-aligned-parsing
on “Sampling and structured outputs in LLMs”Submitted Prompts are using 4x more tokens than you need as a story · 7 points
HN discussionWe’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. https://boundaryml.com/
Comment linking /
on “Sampling and structured outputs in LLMs”Submitted Beating OpenAI's structured outputs on cost, accuracy and speed as a story · 4 points
HN discussionMeanwhile https://dev.to/rishabdugar/crafting-structured-json-response... and https://www.boundaryml.com/blog/structured-output-from-llms You are confusing API response payloads with Structured JSON that we expect to conform to the given Schema. It's carnage that requires defensive coding. Neither OpenAI nor Google are interested in fixing this, because some developers decide to retry until they get valid structured…
Comment linking /blog/structured-output-from-llms
on “MCP Specification – version 2025-06-18 changes”Submitted Every Way to Get Structured Output from LLMs as a story · 4 points
HN discussionI expected to read about the methods used by the libraries to get the structured output and not a comparison of the language compatibility for each. Fortunately the same author have a blog post ( https://www.boundaryml.com/blog/type-definition-prompting-ba... ) explaining how their approach works and how it compares to instructor ( https://github.com/jxnl/instructor ). Basically these libraries provide two things: 1…
Comment linking /blog/type-definition-prompting-baml
on “Every Way to Get Structured Output from LLMs”Submitted BAML is Basically A Made-up Language as a story · 3 points
HN discussion(author here) To be more specific, here's a benchmark that we ran last year, where we compared schema-aligned parsing against constrained decoding (then called "Function Calling (Strict)", the orange ƒ): https://boundaryml.com/blog/sota-function-calling
Comment linking /blog/sota-function-calling
on “Structured outputs create false confidence”Submitted Using UUIDs in prompts is bad as a story · 3 points
HN discussion(on of the creators of BAML here) yep! exactly! that workaround we've found works quite well, but the problem is that its not sufficient to just retry in the case of failed schema matches (its both inefficient and also imo incorrect). Take these two scenarios for example: Scenario 1. My system is designed to output receipts, but the user does something malicious and gives me an invoice. during step 2, it fails to fi…
Comment linking /blog/schema-aligned-parsing
on “Structured outputs create false confidence”Submitted Lambda the Ultimate AI Agent as a story · 3 points
HN discussionhttps://boundaryml.com/ LLM to types and done
Comment linking /
on “Show HN: I built a local-first daily planner for iOS”Submitted Prompts are using 4x more tokens than you need as a story · 3 points
HN discussionThe use case in the article is relatively simple. For more complex structures, BAML ( https://www.boundaryml.com/ ) is a better option.
Comment linking /
on “Structured Output with LangChain and Llamafile”Submitted 5 Years, 12 Pivots as a story · 2 points
HN discussionI kind of like BAML https://boundaryml.com/ been using it in production Edit, read the article -its really good- that cycle of AI engineering progression is spot on -read the article too!
Comment linking /
on “If Dspy is so great, why isn't anyone using it?”Submitted Structured Outputs Create False Confidence as a story · 2 points
HN discussionBAML ( https://boundaryml.com/ ) is really great for handling structured outputs. It's what we use at my company.
Comment linking /
on “LLM Structured Outputs Handbook”Submitted AI Agents Need a New Syntax as a story · 2 points
HN discussion
