Anthropic Claude 4.6
Hacker ApprovedRanked 22 of 72 language models by recommenders
People recommend it for coding, planning and troubleshooting, and many find it reliable when steered well. Critics say it overthinks and goes off the rails in large contexts.
People recommend it for coding, planning and troubleshooting, and many find it reliable when steered well. Critics say it overthinks and goes off the rails in large contexts.
69 people gave an opinion
- recommend it
- 48
- criticize it
- 21
- approval share
- 70%
Recommended in 9 of 13 months
- November 2025: not recommended
- December 2025: recommended
- January 2026: not recommended
- February 2026: recommended
- March 2026: recommended
- April 2026: recommended
- May 2026: recommended
- June 2026: recommended
- July 2026: recommended
- August 2026: recommended
- September 2026: recommended
- October 2026: not recommended
- November 2026: not recommended (month in progress)
Why it carries the badge
- 48 people recommend itNeeds 10 or more, met.
- 2.3 recommenders for every criticNeeds 2 to 1 or better, met.
- Recommended in 9 of 13 monthsNeeds 4 or more, met.
The same three rules apply to every product, and the badge can’t be bought. How the badge works
What people say about it
The reasons people give, with the number of comments that give each one. Open a reason to read what they said.
Why people recommend it
Each mark is one comment
Strong at writing code
“Claude 4.6 is so much better at writing code that nobody cares what OpenAI is doing”
Comment from 1 Mar 2026 in “"We do not think Anthropic should be designated as a supply chain risk"” “4.6 seems like it can properly one/two shot it now without my attention”
Comment from 20 Feb 2026 in “I found a useful Git one liner buried in leaked CIA developer docs” “SWEs using 4.6 barely ever write code anymore at my big tech.”
Comment from 28 Mar 2026 in “Leaked Anthropic Model Presents 'Unprecedented Cybersecurity Risks'” - Comment from 15 Feb 2026 in “IBM Triples Entry Level Job Openings. Finds Limits to AI”
Good enough for everyday tasks
“it's my "good enough" model. im more than pleased at what i can whip up with 4.6”
Comment from 24 Jul 2026 in “Claude Opus 5” “I've been perfectly fine using OpenClaw and Claude 4.6 for the last six months”
Comment from 28 Aug 2026 in “Small Models Have Arrived” “For my day to day tasks 4.6 feels sufficient.”
Comment from 29 May 2026 in “Claude Opus 4.8” Fast at troubleshooting and planning
“I worked with 4.6 and found some improvements for better planning and sustained us”
Comment from 30 Apr 2026 in “Claude.ai and API Unavailable” “The real win is the speed at which Claude troubleshoots, though.”
Comment from 16 Apr 2026 in “Ask HN: Is Claude Getting Worse?” Reliable with steering and freedom
“4.6 is better if you give it more freedom doing it's tasks.”
Comment from 23 May 2026 in “Microsoft Drops Claude Code After Budget Overrun” “as long as we steer it IT does a good enough job”
Comment from 10 Jun 2026 in “Claude Fable 5”
What people warn about
Each mark is one comment
Overthinks and overengineers tasks
“The other day 4.6 was fantastic for x task. Today, 4.6 overengineered everything”
Comment from 9 Jun 2026 in “Claude Fable 5” “4.6 was so slow and often “over thinking” or “over analyzing” the problem space”
Comment from 8 Feb 2026 in “Speed up responses with fast mode” Quality regression after updates
“This was on 4.6 when there was that bad (user-reported) regression in ~March of this year.”
Comment from 22 Jun 2026 in “A Theory of Why Prompt Injection Works” “I have had the same experiences you've had with 4.6 and it was ever since they brought out 4.7.”
Comment from 22 Jun 2026 on Hacker News
Other language models people recommend
The four most recommended other language models, with the same counts for each, so you can compare like for like. The tick on each bar marks the 2 to 1 line the badge needs.
| Product | Recommend | Criticize | Approval | Months | Owned | Badge |
|---|---|---|---|---|---|---|
| Claude Opus | 3,331 | 2,746 | 55% | 13 | 4 mo156 owners | — |
| OpenAI GPT | 1,205 | 987 | 55% | 13 | 6 mo47 owners | — |
| Claude Sonnet | 854 | 586 | 59% | 13 | 6 mo46 owners | — |
| DeepSeek V4 Flash | 828 | 223 | 79% | 8 | 1 mo24 owners | Approved |
| Anthropic Claude 4.6This page | 48 | 21 | 70% | 9 | — | Approved |
What people said
40 of the 71 comments that mention it, newest first. Each links to the original comment.
“feels so much simpler and smarter than all new models”
“I've been perfectly fine using OpenClaw and Claude 4.6 for the last six months”
“It was impressive then and it's still impressive.”
“Claude 4.6 didn’t get it. A higher level model might have, but 4.6 is the last model that doesn’t emit erratic content policy rejections for this task.”
“it's my "good enough" model. im more than pleased at what i can whip up with 4.6”
“I found 4.6 more amenable than 4.8 to style directions”
“I liked having that available, and still use it sometimes.”
“This was on 4.6 when there was that bad (user-reported) regression in ~March of this year.”
“Before that, 4.6 had been a pretty seamless experience.”
“I have had the same experiences you've had with 4.6 and it was ever since they brought out 4.7.”
“as long as we steer it IT does a good enough job”
“I've yet to use a model newer than 4.6-1M-XHIGH that I trust to earn a higher ROI on that workflow”
“The other day 4.6 was fantastic for x task. Today, 4.6 overengineered everything”
“For my day to day tasks 4.6 feels sufficient.”
“performs worse than 4.6 on almost every task”
“4.6 is what changed my mind on LLMs as a tool”
“4.6 is better if you give it more freedom doing it's tasks.”
“4.6 has been excellent when used through the API.”
“I worked with 4.6 and found some improvements for better planning and sustained us”
“so it performs much better in this benchmark and I prefer it”
“the original 4.6 was doing just fine with medium and high”
“I’ll stay on 4.6 for awhile. Seems to be better.”
“felt like a nice knowledgeable coworker who shared his thinking while solving problems”
“the AI will optimize to the point that it breaks other existing functional requirements in the same codebase”
“The real win is the speed at which Claude troubleshoots, though.”
“SWEs using 4.6 barely ever write code anymore at my big tech.”
“it is capable of generating proofs for TLA+ (in TLAPS) (Claude 4.6) with a high success rate”
“made the leap that Claude's 4.6 and GPT's 5.3/5.4 series made”
“These LLMs all go off the rails when the context window gets large.”
“now it usually converges in one or two loops”
“Claude 4.6 is at least a bit resilient to prompt injection”
“Claude 4.6 is so much better at writing code that nobody cares what OpenAI is doing”
“4.6 seems like it can properly one/two shot it now without my attention”
“Claude 4.6 has been excellent with Go, and truly incompetent with Elixir”
“I'm using the 1M context 4.6 and it's great.”
“I'm finding medium effort on 4.6 to give more satisfactory results for that kind of thing.”
“they can do stuff that I've seen senior devs fail at”
“4.6 was so slow and often “over thinking” or “over analyzing” the problem space”
“it took 5 minutes to make something that actually works”
“Still not that recent, but not as bad as GPT.”
Counts come from 1,795,640 Hacker News comments, November 2025 to November 2026. How we count. Report a mistake.