Claude Haiku
Anthropic's smallest, fastest and cheapest Claude large language model.
Recommended as a cheap, fast model for subagents and simple, clearly specified tasks12345, and for turning natural language into structured JSON.678 The most common complaint is that it falls short on serious or complex coding work910111213, and some say newer versions got slower.1415
Why people recommend it
Subagent workhorse · 3
Often paired with Opus or Sonnet as the cheap model for delegated grunt work, tests and explorer subagents.
“Haiku 5.5 is decent, cheap, and fast.”
“I would mostly use Haiku in task or explorer subagents.”
Show 1 more
Cheap and fast · 3
Described as cheap and fast, and as good as Sonnet for many simple work tasks at lower cost.
“Haiku 5.5 is decent, cheap, and fast.”
“at work Haiku almost always works just as well as Sonnet does and its much cheaper/faster”
Show 1 more
“Tweaks and refinements use Haiku. Users can't tell the difference for small edits, and it cut iteration costs ~80%.”
Simple tasks · 3
Used for quick, well-defined jobs like commit messages, log parsing, tagging and latency-sensitive tasks.
“I use haiku for things that needs to be quick, have really clear instructions.”
“At work we use haiku 4.5 for a handful of latency sensitive tasks that are fairly simple. It performs well.”
Show 1 more
“then I use Haiku to do some categorization and auto tagging”
Structured output · 3
Reported as highly accurate at converting natural language into complex structured specs and JSON.
“We use Haiku 4.5 to turn users' natural language queries and requests into fairly complex structured specs for interior design and construction. It has near perfect accuracy.”
“It continues to be absurdly capable for converting natural language to structured JSON”
Show 1 more
“I switched from structured outputs on OpenAI apis to unstructured on Claude (haiku 4.5) and haven't had any issues (yet)”
Surprisingly capable · 3
Some users were surprised by how well it handled harder work and say it keeps getting smarter.
“I used Haiku with Claude Code during the outage, and was surprised at how well it did.”
“Now even Haiku at low effort does.”
Show 1 more
“just letting Haiku rip. Honestly, it blows me away.”
Caveats and dissent
Complex coding · 3
Several users say it struggles with serious coding and falls apart when given larger tasks.
“They are cute but for serious coding they tend to waste your expensive time.”
“any time I ask Opus to "try" letting Haiku do the work, it just falls apart”
Show 1 more
“when I use them via JetBrains AI it feels like only Opus is good enough”
Newer versions slower · 2
Some found the 5.5 release slower and worse than before, including prompt leakage.
“Results so far are trash. Prompt leakage even. And it’s slower.”
“I used to use haiku for the speed. Now its just as slow as the rest.”
Price/performance · 3
Some see cheaper competitors like DeepSeek or local Qwen models as better value.
“Haiku also doesn't seem like a competitive price/performance mix.”
Show 1 more
“Absolute fucking L from anthropic for not introducing a newer cheaper Haiku4.5.”
Subagent overhead · 1
One user found Haiku subagents took too many steps and wrote too much documentation to be worth it.
“In practice the subagents took so many steps and wrote so much documentation that it became not worth it.”
Complex structures · 1
One user says it struggles with complex structured outputs.
“at work Haiku almost always works just as well as Sonnet does and its much cheaper/faster”
Often compared with
1 moved from Claude Haiku to Claude Sonnet
1 moved from GPT Luna to Claude Haiku
1 moved from Claude Haiku to GPT Luna
1 moved from Claude Haiku to Qwen 3.6
Who recommends it in their own words
Threads where it came up
- Claude Haiku 5.5 9 mentions
- Haiku R1/beta6 has been released 5 mentions
- Claude Opus 5.5 3 mentions
- Ask HN: What default model do you use and why? 2 mentions
- Haiku 2 mentions
- Semantic search across major museums 1 mention
Also written as Claude Haiku, Claude Haiku 4.5, Haiku 5.5, Claude Haiku 5.5. Data: claude-haiku.json. Quotes are verbatim; the verdict and theme labels are written by a model from these comments, and every clause links to its sources.
All comments 44
“then I use Haiku to do some categorization and auto tagging”
“the most impressive thing about it to me was that it seemed not to overthink when it wasn't necessary”
“I should try Haiku 5.5 high/max because it's posting higher numbers than Deepseek V4.1 flash for half the price”
“Coding is handled by Haiku 5.5/GPT 6 Luna/DeepSeek v4.1 Flash.”
“Haiku 5.5 is decent, cheap, and fast.”
“Claude Haiku 5.5 and GPT-6 Luna - both excellent inexpensive models”
“I've moved back to Haiku as my daily driver for a lot of stuff.”
“Now even Haiku at low effort does.”
“Haiku also doesn't seem like a competitive price/performance mix.”
“it doesn't quite matter to me if the cheap anthropic model is 20-50% more expensive compared to luna”
“now I want to give it a try”
“I would mostly use Haiku in task or explorer subagents.”
“We use Haiku 4.5 to turn users' natural language queries and requests into fairly complex structured specs for interior design and construction. It has near perfect accuracy.”
“At work we use haiku 4.5 for a handful of latency sensitive tasks that are fairly simple. It performs well.”
“Results so far are trash. Prompt leakage even. And it’s slower.”
“If you mean Haiku 4.5 though I agree, that model was a waste of time and money.”
“I use haiku for things that needs to be quick, have really clear instructions.”
“any time I ask Opus to "try" letting Haiku do the work, it just falls apart”
“I run Haiku, FreeBSD, OpenBSD, Ubuntu, and Pop OS, and all of them go like stink and are daily-drivable”
“Haiku for running tests”
“I highly recommend you give Haiku a spin”
“current Haiku is useless”
“I used to use haiku for the speed. Now its just as slow as the rest.”
“Absolute fucking L from anthropic for not introducing a newer cheaper Haiku4.5.”
“Only reason I used Haiku was when I made Opus 5 run Haiku agents”
“It continues to be absurdly capable for converting natural language to structured JSON”
“Haiku 2.5 s”
“To ask for web search, simple commands and research, or passing logs to analyze it's more than enough”
“I have an 'ask' alias in my shell that just uses Haiku. I use it daily for pretty much everything”
“a skill/script combo suitable for Haiku to accomplish the task, and now that task happens at minimal cost”
“I added Haiku because it looks and performs so well”
“I would love to give Haiku a try. Indeed, I've looked at it's code before (it's quite good, or was, last time I looked at it).”
“impressed by the speed and the reasonable availability of software”
“I love the UI style of HaikuOS”
“But honestly Linux felt faster on this machine. The memory usage was about the same”
“Haiku subagents”
“made a simple skill [1] that gets even Claude Haiku to write correct code”
“while Haiku, Opus and Sonnet are slow and create the same outputs”
“at work Haiku almost always works just as well as Sonnet does and its much cheaper/faster”
“I use Haiku for it, hoping that a dumber model will do a better job.”
“They are cute but for serious coding they tend to waste your expensive time.”
“when I use them via JetBrains AI it feels like only Opus is good enough”
“In practice the subagents took so many steps and wrote so much documentation that it became not worth it.”
“I switched from structured outputs on OpenAI apis to unstructured on Claude (haiku 4.5) and haven't had any issues (yet)”