Grok
xAI's large language model and chatbot assistant, also integrated into X and Cursor.
HN commenters like Grok for fast, cheap coding and code review1234 and for direct, well-sourced research answers.567 The most common complaints are distrust of the company's data handling and conduct89101112, plus confident answers that are often wrong.131415
Why people recommend it
Speed · 3
Many say it is much faster than Claude, which keeps coding sessions interactive.
“Grok is 3x+ faster than Claude and I can't tell the diff in engineering work quality.”
“This is part of why I switched to Grok 4.5”
Show 1 more
“I hopped over to Grok4.6 after the announcement the other day and its been pretty good so far. It also is very fast.”
Price · 3
Seen as cheap for its capability, by API and through Cursor or SuperGrok.
“I've now switched to Grok 4.6, I was previously using Opus 4.8 but Grok is good enough now, and way cheaper”
“"opus level intelligence for haiku prices" which is a really big deal for application developers”
Show 1 more
“it's quite a good deal from the Cursor subscription side but it's cheap even by API prices.”
Coding · 3
Considered a capable implementation model, some say near Opus level.
“In my experience Grok 4.5 codes at Opus 4.8 level, and being much faster as cheaper”
“I find it to be a perfectly capable model for implementation”
Show 1 more
“I use Grok 4.5 High Fast in Cursor (Agentic window), and it's a genuinely great model.”
Code review · 3
Used as a reviewer that finds bugs other models missed.
“$20 a month each for Grok and GLM have easily paid for themselves by finding bugs that my main work models didn’t”
“Or try Grok, 4.5 seems pretty capable, should be close to K3 in many coding tasks.”
Show 1 more
“I use Sol and Grok 4.5 as my inline debuggers/reviewers, and both do well”
Research · 3
Praised for research mode and searching and referencing sources.
“Grok is my go-to LLM for research not because I care for Elon Musk but because it's really good.”
“grok does good research and produces less bullshit, especially when prompted to be critical of an idea”
Show 1 more
Direct style · 3
Users like its concise, direct, less personable responses.
“It communicates better.”
“Grok is indeed fact/truth oriented, direct, and less personable (which I vastly prefer).”
Show 1 more
“Recently added Grok and it has somehow bacome my second most used model.”
Caveats and dissent
Company trust · 3
Some reject it over the owner's conduct and platform controversies.
“When people started using Grok to make CSAM the platform's response was to paywall image generation.”
Show 1 more
“US military leadership switched to Grok, whose leader had no such qualms”
Data privacy · 3
Reports of Grok Build uploading code bases and an organizational ban over data concerns.
“if you like grok-4.5 model (it is a good model)”
Show 1 more
“My organization has outright banned Grok, because we don't trust SpaceX to hold up to contractual agreements vis-a-vis data-privacy/training.”
Accuracy · 3
Described as confidently incorrect and prone to hallucination.
“Grok is the most confident of any model I use. Unfortunately, it is also the most often incorrect.”
“Grok is subject to hallucinations has been caught making up posts wholesale.”
Show 1 more
“Grok and Gemini both failed in a similar way.”
Weaker than rivals · 3
Some find it behind Opus or useless for tasks like writing critique.
“Grok is capable, but disappointing compared to Opus.”
“Whereas Grok is absolutely useless at it”
Show 1 more
“The code they supplied to me has NEVER worked or was so convoluted that I threw it away and did it myself.”
Bias · 2
Some commenters raise concerns about political slant in its answers.
“AFAIK X published their algorithm which more or less just says: "we use grok to decide which posts to show".”
Speed varies · 2
A few found some versions slow.
“But I am not liking them much. And they are extremely slow as well.”
Often compared with
3 moved from Claude to Grok
4 moved from ChatGPT to Grok
3 moved from Claude Opus to Grok
1 moved from Grok to Claude Opus
1 moved from Grok to Claude Code
Who recommends it in their own words
Threads where it came up
- Grok 4.6 9 mentions
- Grok 4.7 8 mentions
- SpaceXAI's Grok 4.6 Scores 61 on the Artificial Analysis Intelligence Index 6 mentions
- Grok 4.5 6 mentions
- Ask HN: What default model do you use and why? 5 mentions
- Grok outage 3 mentions
Also written as Grok 4.5, Grok 4.6, Grok 4.1. Data: grok.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 52
“I had Grok Bot make me a very basic version of Figma.”
“I asked Grok Bot on another machine if it could get into that one over SSH”
“I also made sure Opus controls Grok in Cursor”
“I don't actually care that much for Grok the model because I think they are more or less homogenous”
“Grok is the most confident of any model I use. Unfortunately, it is also the most often incorrect.”
“AFAIK X published their algorithm which more or less just says: "we use grok to decide which posts to show".”
“I have Pro max subs for Claude, Codex, Grok, and Google AI”
“grok bot in the cloud on everything else”
“$20 a month each for Grok and GLM have easily paid for themselves by finding bugs that my main work models didn’t”
“Grok bot has recently been released; it has its own limits and also helps save Cursor tokens”
“Grok whilst did solid work, struggled with cohesive architecture”
“The code they supplied to me has NEVER worked or was so convoluted that I threw it away and did it myself.”
“I use the crap out of Grok in my daily life”
“I've found Grok in vehicle to have degraded.”
“I used it a year or so ago when it came out and it seemed as gated as all the others.”
“It’s the consumer app and model I like the most, but my sentiment has changed.”
“4.6 was a slower less enjoyable version of that.”
“I use it in the car to talk to because it's built-in, mostly for navigation via voice.”
“Recently added Grok and it has somehow bacome my second most used model.”
“I find it to be a perfectly capable model for implementation”
“grok does good research and produces less bullshit, especially when prompted to be critical of an idea”
“They are great help and provide me invaluable advice.”
“But I am not liking them much. And they are extremely slow as well.”
“I use Claude, Codex, Grok and Muse for all the other projects”
“Four AI models (ChatGPT, Claude, Grok, Gemini) each write a trading rulebook”
“Grok is keeping up and really good at explicit stuff”
“I’m increasingly turning to Grok here.”
“Grok is capable, but disappointing compared to Opus.”
“the latest Grok is actually a beast of a model, but expensive enough to not be a stand out.”
“Esp if that other model is as unhinged as Grok.”
“Grok is my go-to LLM for research not because I care for Elon Musk but because it's really good.”
“It was 100% vibe-coded using Claude and some Grok”
“Grok is worse, Claude is also worse.”
“I think the one that surprises me the most often is probably grok, but I wouldn't want grok to be my daily driver.”
“It's my favorite chat box for general use.”
“recent versions are good for writing specs, code review, etc.”
“I like grok 4.6 more than any anthropic or claude model.”
“With grok bot, I've just been giving it a time to stop working; otherwise, it's likely to burn through my entire week's usage in an afternoon.”
“auto picks between Opus 5, GPT 5.6, Grok 4.6, and Composer 2.5”
“I'm now using Grok 4.6 most of the time.”
“grok uploads your entire code base”
“Grok and Gemini both failed in a similar way.”
“I've now switched to Grok 4.6, I was previously using Opus 4.8 but Grok is good enough now, and way cheaper”
“Grok is indeed fact/truth oriented, direct, and less personable (which I vastly prefer).”
“Grok is 3x+ faster than Claude and I can't tell the diff in engineering work quality.”
“Literal trash for nazis.”
“My organization has outright banned Grok, because we don't trust SpaceX to hold up to contractual agreements vis-a-vis data-privacy/training.”
“Grok is subject to hallucinations has been caught making up posts wholesale.”
“When people started using Grok to make CSAM the platform's response was to paywall image generation.”
“Or try Grok, 4.5 seems pretty capable, should be close to K3 in many coding tasks.”
“if you like grok-4.5 model (it is a good model)”
“So I use Grok Build with 4.5 as its VERY fast and cheap”