Hacker ApprovedDepartments

Gemini Deep Research

Ranked 2 of 6 research agents by recommenders

People recommend it for long, source-backed reports that reach beyond ordinary search. Critics most often say its reports are verbose and contain fabricated content.

People recommend it for long, source-backed reports that reach beyond ordinary search. Critics most often say its reports are verbose and contain fabricated content.

46 people gave an opinion

recommend it
27
criticize it
19
approval share
59%
Each figure is one person. RecommendsCriticizes

Recommended in 9 of 13 months

  1. November 2025: recommended
  2. December 2025: recommended
  3. January 2026: recommended
  4. February 2026: recommended
  5. March 2026: recommended
  6. April 2026: recommended
  7. May 2026: recommended
  8. June 2026: recommended
  9. July 2026: not recommended
  10. August 2026: recommended
  11. September 2026: not recommended
  12. October 2026: not recommended
  13. November 2026: not recommended (month in progress)

Why it doesn’t carry the badge

  • 27 people recommend itNeeds 10 or more, met.
  • 1.4 recommenders for every criticNeeds 2 to 1 or better, not 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

  • Gathers many sources with backlinks

    “Deep Research reports I think are above average internet quality, they collect hundreds of sources, synthesize and contrast them & provide backlinks.”

    Comment from 1 Dec 2025 in “Search tool that only returns content created before ChatGPT's public release”

    “especially if you use Deep Research to ground it in multiple sources”

    Comment from 6 Nov 2025 in “The trust collapse: Infinite AI content is awful”
  • Produces detailed, in-depth reports

    “today I've already had Gemini Deep Research write investigative reports previously unfeasible for me”

    Comment from 17 May 2026 in “Accelerando (2005)”

    “The results of Gemini go deep, are thoughtful and make sense.”

    Comment from 13 Nov 2025 in “GPT-5.1: A smarter, more conversational ChatGPT”

    “I used Gemini Deep Research to build a high-fidelity reference of Kannada linguistics”

    Comment from 2 Mar 2026 in “Show HN: Swalpa – Spoken Kannada learning platform built with Antigravity”
  • Surfaces information beyond normal search

    “you can consume information that's more out of distribution than what you would have consumed had you done only Google Searches”

    Comment from 19 Feb 2026 in “AI makes you boring”

    “Deep research at least helps with dowsing for the knowledge”

    Comment from 24 Mar 2026 in “Is anybody else bored of talking about AI?”
  • Helpful for planning and research

    “I love it. Most recently was reviewing PhD programs in my area then deep diving into faculty research areas.”

    Comment from 3 Nov 2025 in “Tongyi DeepResearch – open-source 30B MoE Model that rivals OpenAI DeepResearch”

    “was able to give me a well planned 4 day trip to a major city”

    Comment from 9 Jun 2026 in “Siri AI”

What people warn about

Each mark is one comment

  • Fabricates or hallucinates content

    “Gemini did the same thing but the report was about twenty pages. 100% hallucinated.”

    Comment from 4 Jul 2026 in “Gone but Not Forgotten: Recovering the Dead Web”

    “If I question the AI on the report, they both admit they just made it up.”

    Comment from 20 Feb 2026 in “Gemini 3.1 Pro Preview”

    “it just lied and fabricated the result of a theorem in its output”

    Comment from 26 Jan 2026 in “Case study: Creative math – How AI fakes proofs”
  • Verbose with long academic preamble

    “its so overly verbose, with no easy way to tone down the wordiness.”

    Comment from 4 Feb 2026 in “Microsoft's Pivotal AI Product Is Running into Big Problems”

    “Nothing I can do prevents it from giving me a page of background that tries to make the question sound important and interesting, even when it isn't.”

    Comment from 8 Sep 2026 in “I-have-ADHD: A skill to stop coding agents from burying the answer”

    “It always starts with a long grandiose preamble and then sounds way too academic”

    Comment from 17 Sep 2026 in “I had Gemini train its own replacement for $9”

Other research agents people recommend

The four most recommended other research agents, 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.

Gemini Deep Research compared with 4 alternatives
ProductRecommendCriticizeApprovalMonthsOwnedBadge
Autoresearch by Karpathy371571%6—Approved
Gemini Deep ResearchThis page271959%9——
OpenAI Deep Research131546%8——
Kagi Research90100%7——
LangAlpha8280%1——
See all 6 research agents, ranked

What people said

40 of the 54 comments that mention it, newest first. Each links to the original comment.

“It always starts with a long grandiose preamble and then sounds way too academic”

CriticizesComment from 17 Sep 2026 in “I had Gemini train its own replacement for $9”

“it basically said it was a bunch of "hallucinated bullshit"”

CriticizesComment from 17 Sep 2026 in “I had Gemini train its own replacement for $9”

“Nothing I can do prevents it from giving me a page of background that tries to make the question sound important and interesting, even when it isn't.”

CriticizesComment from 8 Sep 2026 in “I-have-ADHD: A skill to stop coding agents from burying the answer”

“Gemini's Deep Research is pretty good.”

RecommendsComment from 16 Aug 2026 in “ChatGPT lost 22 points of web share in a year”

“Deep Research gives noticeably better results.”

RecommendsComment from 14 Aug 2026 in “Introducing Toast 1”

“holy crap I had literally just lived a SciFi scene!”

RecommendsComment from 1 Aug 2026 in “Charlie Stross – On the non-use of AI in my writing process”

“Gemini did the same thing but the report was about twenty pages. 100% hallucinated.”

CriticizesComment from 4 Jul 2026 in “Gone but Not Forgotten: Recovering the Dead Web”

“Google has their own "Deep Research" option in this space which seems to work well.”

RecommendsComment from 10 Jun 2026 in “DiffusionGemma: 4x Faster Text Generation”

“was able to give me a well planned 4 day trip to a major city”

RecommendsComment from 9 Jun 2026 in “Siri AI”

“Gemini and Gemini deep research are awesome. The export to google docs in Gemini deep research is tough to beat too.”

RecommendsComment from 20 May 2026 in “An OpenAI model has disproved a central conjecture in discrete geometry”

“today I've already had Gemini Deep Research write investigative reports previously unfeasible for me”

RecommendsComment from 17 May 2026 in “Accelerando (2005)”

“This will unlock a whole new level of capability.”

RecommendsComment from 8 Apr 2026 in “Show HN: Gemma 4 Multimodal Fine-Tuner for Apple Silicon”

“In a typical Gemini Deep Research fashion, it came back with a very obscure recommendation claiming everyone is using it.”

CriticizesComment from 6 Apr 2026 in “Self-hosted disposable sandboxes with exe.dev-like UX”

“I felt like OpenAI's other products, like Deep Research, agent mode, etc, were the same way.”

RecommendsComment from 25 Mar 2026 in “OpenAI set to discontinue Sora video platform”

“Deep research at least helps with dowsing for the knowledge”

RecommendsComment from 24 Mar 2026 in “Is anybody else bored of talking about AI?”

“The result is pretty good!”

RecommendsComment from 22 Mar 2026 in “Five Years of Running a Systems Reading Group at Microsoft”

“But how do I know Gemini didn't hallucinate the same things Claude did?”

MixedComment from 7 Mar 2026 in “Verification debt: the hidden cost of AI-generated code”

“ask for 5-10 researches focused on details and the majority will be wasted on the introduction and already explored topics”

MixedComment from 4 Mar 2026 in “Show HN: Skill for structured deep research with Claude Code and Obsidian”

“I used Gemini Deep Research to build a high-fidelity reference of Kannada linguistics”

RecommendsComment from 2 Mar 2026 in “Show HN: Swalpa – Spoken Kannada learning platform built with Antigravity”

“asking for the why makes for a much better response”

RecommendsComment from 1 Mar 2026 in “AI is making junior devs useless”

“GDR is one of the best debugging tools for problems where you don't have full understanding.”

RecommendsComment from 27 Feb 2026 in “We gave terabytes of CI logs to an LLM”

“OpenAI and Google's Deep Research produce a very long, 100% made up report.”

CriticizesComment from 20 Feb 2026 in “Gemini 3.1 Pro Preview”

“If I question the AI on the report, they both admit they just made it up.”

CriticizesComment from 20 Feb 2026 in “Gemini 3.1 Pro Preview”

“you can consume information that's more out of distribution than what you would have consumed had you done only Google Searches”

RecommendsComment from 19 Feb 2026 in “AI makes you boring”

“their Deep Research usually works the hardest, as long as I run it during off-hours”

RecommendsComment from 13 Feb 2026 in “Gemini 3 Deep Think”

“for me the results where too technical and not structural or high level as i describe them here”

CriticizesComment from 6 Feb 2026 in “How to effectively write quality code with AI”

“its so overly verbose, with no easy way to tone down the wordiness.”

MixedComment from 4 Feb 2026 in “Microsoft's Pivotal AI Product Is Running into Big Problems”

“I think tools like Gemini deep research and NoteBookLM are superb tools.”

RecommendsComment from 1 Feb 2026 in “Show HN: Moltbook – A social network for moltbots (clawdbots) to hang out”

“it just lied and fabricated the result of a theorem in its output”

CriticizesComment from 26 Jan 2026 in “Case study: Creative math – How AI fakes proofs”

“the most consistent at actually using the search results for this”

RecommendsComment from 22 Jan 2026 in “Satya Nadella: "We need to find something useful for AI"”

“Claude code and Gemini deep research is much much more better in terms of quality”

RecommendsComment from 31 Dec 2025 in “OpenAI's cash burn will be one of the big bubble questions of 2026”

“a clear shift from simple search augmentation toward fully autonomous, research-grade intelligence”

RecommendsComment from 13 Dec 2025 on Hacker News

“Try the same question in deep research mode.”

RecommendsComment from 12 Dec 2025 in “GPT-5.2”

“Deep Research reports I think are above average internet quality, they collect hundreds of sources, synthesize and contrast them & provide backlinks.”

RecommendsComment from 1 Dec 2025 in “Search tool that only returns content created before ChatGPT's public release”

“after ten minutes of trawling through obscure forum posts it came back with a blunt "No"”

CriticizesComment from 23 Nov 2025 in “Google Revisits JPEG XL in Chromium After Earlier Removal”

“Because the $250 per-month price for Gemini's deep-research tier is hard to justify right now”

MixedComment from 13 Nov 2025 in “GPT-5.1: A smarter, more conversational ChatGPT”

“The results of Gemini go deep, are thoughtful and make sense.”

RecommendsComment from 13 Nov 2025 in “GPT-5.1: A smarter, more conversational ChatGPT”

“especially if you use Deep Research to ground it in multiple sources”

RecommendsComment from 6 Nov 2025 in “The trust collapse: Infinite AI content is awful”

“I like my coding assistant and deep.research”

RecommendsComment from 5 Nov 2025 in “Don't panic yet, investors say as high-flying AI stocks tumble”

“I love it. Most recently was reviewing PhD programs in my area then deep diving into faculty research areas.”

RecommendsComment from 3 Nov 2025 in “Tongyi DeepResearch – open-source 30B MoE Model that rivals OpenAI DeepResearch”

Counts come from 1,795,640 Hacker News comments, November 2025 to November 2026. How we count. Report a mistake.