Hacker ApprovedDepartments

Unsloth Studio

Hacker Approved

Ranked 5 of 27 local LLM runners by recommenders

An open-source local AI app with web search and tool calling, from the makers of fast-arriving, high-quality quantized models, though installation and Python setup are hard, especially on macOS.

An open-source local AI app with web search and tool calling, from the makers of fast-arriving, high-quality quantized models, though installation and Python setup are hard, especially on macOS.

121 people gave an opinion

recommend it
94
criticize it
27
approval share
78%
Each figure is one person. RecommendsCriticizes

Recommended in 11 of 13 months

  1. November 2025: recommended
  2. December 2025: not 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: recommended
  10. August 2026: recommended
  11. September 2026: recommended
  12. October 2026: recommended
  13. November 2026: not recommended (month in progress)

Why it carries the badge

  • 94 people recommend itNeeds 10 or more, met.
  • 3.5 recommenders for every criticNeeds 2 to 1 or better, met.
  • Recommended in 11 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

  • Good quants of released models

    “open source alternative to the more widely known LM Studio, and also built by the people who make good quantizations of released models”

    Comment from 21 May 2026 in “Indexing a year of video locally on a 2021 MacBook with Gemma4-31B (50GB swap)”

    “They provide excellent documentation and they're often very quick to get high quality quants up in major formats.”

    Comment from 20 Feb 2026 in “Ggml.ai joins Hugging Face to ensure the long-term progress of Local AI”

    “Unsloth Studio is open source, run by the same Unsloth that produces some of the best quants in the business”

    Comment from 16 Jul 2026 in “LM Studio Bionic: the AI agent for open models”
  • Open source local AI tooling

    “open source alternative to the more widely known LM Studio, and also built by the people who make good quantizations of released models”

    Comment from 21 May 2026 in “Indexing a year of video locally on a 2021 MacBook with Gemma4-31B (50GB swap)”

    “Unsloth Studio is open source, run by the same Unsloth that produces some of the best quants in the business”

    Comment from 16 Jul 2026 in “LM Studio Bionic: the AI agent for open models”

    “it's actually open source and I trust Unsloth via their quantized models a lot more than LM Studio”

    Comment from 16 Jul 2026 in “LM Studio Bionic: the AI agent for open models”
  • Excellent documentation and guides

    “They are very active in the local model community and have extensive documentation and tooling to help with running and training models locally.”

    Comment from 12 Aug 2026 in “Qwen/Qwen3.8-2.4T-A95B”

    “They provide excellent documentation and they're often very quick to get high quality quants up in major formats.”

    Comment from 20 Feb 2026 in “Ggml.ai joins Hugging Face to ensure the long-term progress of Local AI”

    “has great guides and provides everything you need to get started on Google Colab for free”

    Comment from 20 Feb 2026 in “Ggml.ai joins Hugging Face to ensure the long-term progress of Local AI”
  • Web search and tool calling

    “it has web search and self-healing tool calls so having a web-searching local ChatGPT alternative is very easy to spin up”

    Comment from 22 May 2026 in “Indexing a year of video locally on a 2021 MacBook with Gemma4-31B (50GB swap)”

    “Unsloth Studio is more featureful (well integrated tool calling, web search, and code execution being headline features)”

    Comment from 3 Apr 2026 in “April 2026 TLDR Setup for Ollama and Gemma 4 26B on a Mac mini”

    “I'm using the 2B model for web search and structured JSON output back via Unsloth Studio and its API”

    Comment from 5 Jun 2026 in “Gemma 4 QAT models: Optimizing compression for mobile and laptop efficiency”

What people warn about

Each mark is one comment

  • Difficult installation and Python setup

    “still have trouble to get unsloth studio setup working on my MacBook because of Python tooling issues”

    Comment from 18 Mar 2026 in “Unsloth Studio”

    “Unsloth looks great, wanted to check it out, but there is no app file to download and install?”

    Comment from 19 Mar 2026 in “Unsloth Studio”

    “It seems we might need more instructions on how to set up python (via uv) in vanilla MacOS.”

    Comment from 18 Mar 2026 in “Unsloth Studio”
  • Gaps in model and feature support

    “Unsloth Studio which is an OSS runner I use is still in progress with MLX support although it's still a ways away”

    Comment from 20 Jul 2026 in “Nativ: Run frontier open models locally on your Mac”

    “when I searched for gemma-4 in Unsloth Studio it only shows Gemma 3 models”

    Comment from 2 Apr 2026 in “Google releases Gemma 4 open models”

    “It's a pity that even Unsloth gave up low-VRAM finetuning with MoE”

    Comment from 27 Jul 2026 in “Kimi-K3 Releases on HuggingFace 6/27”
  • Rough downloads and frontend

    “the downloading functionality is half-baked and didn’t cope with resuming downloads”

    Comment from 21 May 2026 in “Indexing a year of video locally on a 2021 MacBook with Gemma4-31B (50GB swap)”

    “feels like a vibe-coded frontend based on already given backend features”

    Comment from 17 Mar 2026 in “Unsloth Studio”

    “It really doesn't react well to having you load the coding model”

    Comment from 22 Sep 2026 in “I asked Meta’s Muse for its filesystem and it sent me 6.8GB”
  • Quantization claims questioned

    “The fact that Unsloth only just started publishing KL divergences shows how unserious the quantization space is.”

    Comment from 14 Aug 2026 in “Qwen3.8-27B”

    “unsloth and byteshape are just using and highlighting features that have been available the whole time”

    Comment from 17 Apr 2026 in “Qwen3.6-35B-A3B: Agentic Coding Power, Now Open to All”

Other local LLM runners people recommend

The four most recommended other local LLM runners, 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.

Unsloth Studio compared with 4 alternatives
ProductRecommendCriticizeApprovalMonthsOwnedBadge
llama.cpp47110981%131 yr4 ownersApproved
Ollama35024759%121.5 yrs8 owners—
LM Studio2177475%123 yrs4 ownersApproved
DS4 by antirez1092084%71 mo3 ownersApproved
Unsloth StudioThis page942778%11—Approved
See all 27 local LLM runners, ranked

What people said

40 of the 167 comments that mention it, newest first. Each links to the original comment.

“It really doesn't react well to having you load the coding model”

CriticizesComment from 22 Sep 2026 in “I asked Meta’s Muse for its filesystem and it sent me 6.8GB”

“For a GUI experience that also serves an OpenAI-compatible API, Unsloth studio or LM Studio is probably the way to go”

RecommendsComment from 4 Sep 2026 in “Which tools do Claude, Codex and Cursor choose? We measured 17k runs to find out”

“The fact that Unsloth only just started publishing KL divergences shows how unserious the quantization space is.”

CriticizesComment from 14 Aug 2026 in “Qwen3.8-27B”

“They are very active in the local model community and have extensive documentation and tooling to help with running and training models locally.”

RecommendsComment from 12 Aug 2026 in “Qwen/Qwen3.8-2.4T-A95B”

“lm-studio or unsloth are to llama.cpp as something like phpmyadmin is to mysql/mariadb.”

RecommendsComment from 12 Aug 2026 in “llama.cpp”

“Installation failed: create venv (system Python) failed (exit code 2)”

CriticizesComment from 12 Aug 2026 in “Introducing Unsloth Desktop”

“We launched a Desktop app with fast diffusion, video gen, training support, inference + tool calling”

RecommendsComment from 11 Aug 2026 in “Unsloth Desktop: Open-Source App for Local Models”

“It's a pity that even Unsloth gave up low-VRAM finetuning with MoE”

CriticizesComment from 27 Jul 2026 in “Kimi-K3 Releases on HuggingFace 6/27”

“Unsloth Studio which is an OSS runner I use is still in progress with MLX support although it's still a ways away”

MixedComment from 20 Jul 2026 in “Nativ: Run frontier open models locally on your Mac”

“it spewed a bunch of stuff into the system's pip archive and broke other software”

CriticizesComment from 19 Jul 2026 in “Ollama: All Aboard Open Models”

“Unsloth Studio is open source, run by the same Unsloth that produces some of the best quants in the business”

RecommendsComment from 16 Jul 2026 in “LM Studio Bionic: the AI agent for open models”

“it's actually open source and I trust Unsloth via their quantized models a lot more than LM Studio”

RecommendsComment from 16 Jul 2026 in “LM Studio Bionic: the AI agent for open models”

“unsloth has their own studio type app that might be even better suited for these quants”

RecommendsComment from 2 Jul 2026 in “Kimi K2.7 Code is generally available in GitHub Copilot”

“Unsloth make both the Studio and the quants which fix various issues with many models”

RecommendsComment from 30 Jun 2026 in “Qwen 3.6 27B is the sweet spot for local development”

“I am fully impressed by the diligence and quality of some of the open source tooling”

RecommendsComment from 23 Jun 2026 in “OpenAI DayBreak – GPT-5.5-Cyber”

“I'm using the 2B model for web search and structured JSON output back via Unsloth Studio and its API”

RecommendsComment from 5 Jun 2026 in “Gemma 4 QAT models: Optimizing compression for mobile and laptop efficiency”

“it has web search and self-healing tool calls so having a web-searching local ChatGPT alternative is very easy to spin up”

MixedComment from 22 May 2026 in “Indexing a year of video locally on a 2021 MacBook with Gemma4-31B (50GB swap)”

“the downloading functionality is half-baked and didn’t cope with resuming downloads”

CriticizesComment from 21 May 2026 in “Indexing a year of video locally on a 2021 MacBook with Gemma4-31B (50GB swap)”

“open source alternative to the more widely known LM Studio, and also built by the people who make good quantizations of released models”

RecommendsComment from 21 May 2026 in “Indexing a year of video locally on a 2021 MacBook with Gemma4-31B (50GB swap)”

“it provides MTP support now which 2x the token generation speed with no loss of accuracy”

RecommendsComment from 20 May 2026 in “Qwen3.7-Max: The Agent Frontier”

“Unsloth releases a lot of good quantizations and has great support for the latest features”

RecommendsComment from 20 May 2026 in “I Don't Vibe Code”

“It is supposed to be faster + they will update the new models multiple times during the first month to correct bugs and performance issues”

RecommendsComment from 19 May 2026 in “Qwen 3.7 Preview”

“what Han brothers are doing with Unsloth in AI, and stories like this one, makes this fellow Aussie super proud”

RecommendsComment from 16 May 2026 in “How an Australian Teen Team Is Making Radio Astronomy Affordable for Schools”

“running the studio is a bit of a nightmare when it calls diskpart during its install”

CriticizesComment from 22 Apr 2026 in “Qwen3.6-27B: Flagship-Level Coding in a 27B Dense Model”

“unsloth and byteshape are just using and highlighting features that have been available the whole time”

CriticizesComment from 17 Apr 2026 in “Qwen3.6-35B-A3B: Agentic Coding Power, Now Open to All”

“Unsloth Studio is more featureful (well integrated tool calling, web search, and code execution being headline features)”

RecommendsComment from 3 Apr 2026 in “April 2026 TLDR Setup for Ollama and Gemma 4 26B on a Mac mini”

“we recommend specific quants (like Gemma-4) and also auto calculate the context length etc!”

RecommendsComment from 2 Apr 2026 in “Google releases Gemma 4 open models”

“when I searched for gemma-4 in Unsloth Studio it only shows Gemma 3 models”

CriticizesComment from 2 Apr 2026 in “Google releases Gemma 4 open models”

“Since Daniel does the work of ten, it seems like their value:cost ratio is world-class at the very least.”

RecommendsComment from 19 Mar 2026 in “Unsloth Studio”

“Unsloth looks great, wanted to check it out, but there is no app file to download and install?”

MixedComment from 19 Mar 2026 in “Unsloth Studio”

“It seems we might need more instructions on how to set up python (via uv) in vanilla MacOS.”

CriticizesComment from 18 Mar 2026 in “Unsloth Studio”

“still have trouble to get unsloth studio setup working on my MacBook because of Python tooling issues”

CriticizesComment from 18 Mar 2026 in “Unsloth Studio”

“the UI was primarily a culmination of pain points folks had when doing either training or inference”

RecommendsComment from 17 Mar 2026 in “Unsloth Studio”

“feels like a vibe-coded frontend based on already given backend features”

CriticizesComment from 17 Mar 2026 in “Unsloth Studio”

“Happy to see unsloth making it even easier for people like me to get going with fine tuning”

RecommendsComment from 17 Mar 2026 in “Unsloth Studio”

“This time even Unsloth could not provide bitsandbytes 4-bit models”

MixedComment from 5 Mar 2026 in “Qwen3.5 Fine-Tuning Guide – Unsloth Documentation”

“Unsloth's approach of patching the attention kernels at the Python level rather than requiring custom CUDA is what makes this accessible”

RecommendsComment from 4 Mar 2026 in “Qwen3.5 Fine-Tuning Guide – Unsloth Documentation”

“Unsloth is one of, if not the most well-known provider of model quantizations”

RecommendsComment from 2 Mar 2026 in “Qwen3.5 Small: 0.8B, 2B, 4B, 9B Released”

“has great guides and provides everything you need to get started on Google Colab for free”

RecommendsComment from 20 Feb 2026 in “Ggml.ai joins Hugging Face to ensure the long-term progress of Local AI”

“They provide excellent documentation and they're often very quick to get high quality quants up in major formats.”

RecommendsComment from 20 Feb 2026 in “Ggml.ai joins Hugging Face to ensure the long-term progress of Local AI”

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