Model Follower.

Up to date

Rebuilt
2026-10-25 06:54 UTC
Release radar
117 on watch
Hub snapshot
2026-10-25
Re-checked
daily · every number sourced and dated

Models › Qwen-2.5 family

Qwen-2.5-1B-RLCD

Hugging Face ↗ Follow changes

harshatheg Apache 2.0 out 39 days ago

Text in, text out.

Verified 2026-10-25 Rank #17401 in Decision models · score 23.9 method v1 Sources 3 Downloads 30d 0 · #618 of 619 tracked

Can I use it?Apache 2.0Permissive; patent grantlicence file2026-10-24
Do people use it?0downloads, 30 days#618 of 619 we track · below the median+3 likes this week590 likesHugging Face2026-10-24

Worth reading

Picked by rule from 20 tracked pieces: the lab's own post, the most shared, a field report, the latest.

Most discussed · 544 HN pts

Qwen2.5-VL-32B: Smarter and Lighter

Alibaba open-sourced Qwen2.5-VL-32B-Instruct under the Apache 2.0 license, optimizing the model with reinforcement learning to improve human alignment and mathematical reasoning.

  • The model achieves significant advantages in MMMU, MMMU-Pro, and MathVista benchmarks focusing on complex, multi-step reasoning.
  • Qwen2.5-VL-32B-Instruct outperforms its predecessor Qwen2-VL-72B-Instruct by a substantial margin on the MM-MT-Bench benchmark.
qwenlm.github.io 2025-03-24 544 HN pts · 293 comments 606 words about the Qwen 2.5 family family
Latest · 21 months ago

Qwen2.5-Max: Exploring the intelligence of large-scale MoE model

Alibaba Cloud releases Qwen2.5-Max, a large-scale MoE model pretrained on over 20 trillion tokens and post-trained with SFT and RLHF.

qwenlm.github.io 2025-01-28 118 HN pts · 30 comments 556 words about the Qwen 2.5 family family

20 best of 20 tracked

The 2 picks above, plus these 18. Top is the radar's score: discussion, source authority and the lab's own posts, fading with age.

  1. VideoPewDiePie announces his own Qwen 2.5 9B fine-tune [video]youtube.com · 2026-10-01
  2. ArticleQwen2.5-Coder-32B is an LLM that can code well that runs on my Macsimonwillison.net · 2024-11-13 · 164 HN pts
  3. ArticleQwen2.5: A Party of Foundation Modelsqwenlm.github.io · 2024-09-18 · 168 HN pts
  4. CodeShow HN: Qwen-2.5-32B is now the best open source OCR modelgithub.com · 2025-04-01
  5. CodeLocal autonomous security agent powered by Qwen 2.5-7B on Kali Linuxgithub.com · 2026-05-07
  6. CodeQwen2.5-Coder-32B with coding abilities matching those of GPT-4ogithub.com · 2024-11-09 · 21 HN pts
  7. ArticleQwen2.5-Coder Beats GPT-4o and Claude 3.5 Sonnet in Codingqwenlm.github.io · 2024-11-11 · 23 HN pts
  8. ArticleAlibaba Qwen2.5-Omni-7B: Open Source End-to-End Multimodal AI Modelalizila.com · 2025-03-27 · 29 HN pts
  9. ArticleQwen2.5-VLqwenlm.github.io · 2025-01-27 · 8 HN pts
  10. ArticleQwen2.5-VL-32B: Smarter and Lightersimonwillison.net · 2025-03-25 · 3 HN pts
  11. ArticleQwen2.5-Max: The Next Leap in AI Innovationmedium.com · 2025-01-29 · 17 HN pts
  12. CodeRun Qwen 2.5 Family and Other LMs in GGUF Format from HF Locallygithub.com · 2024-09-18
  13. ArticleQwen2.5-Omni Technical Reporthuggingface.co · 2025-03-30
  14. ArticleQwen2.5 Omni: See, Hear, Talk, Write, Do It Allqwenlm.github.io · 2025-03-26
  15. PaperQwen2.5-Coder Technical Reportarxiv.org · 2024-11-11
  16. ArticleShow HN: Using RasPi and Qwen 2.5 VL to find out when birds visit my balconyakshay326.com · 2026-04-04
  17. DiscussionQwen2.5-coder on the events occurring during the Tiananmen Square protestsnews.ycombinator.com · 2024-11-20
  18. ArticleGuide how to use a local LLM Qwen2.5-coder-7B as an autocompelte in VSCodeeamag.me · 2024-10-29

Where it's used

Hugging Face does not say who downloads a model or with what. These are the proxies: API providers listing it, quantized builds, fine-tunes/adapters/merges, and Spaces that load it.

No API provider serves it, so its downloads are people running it themselves: 6 Spaces.

  • As a service0no API provider listed
  • On own hardware0quantized builds
  • Built on00 fine-tunes · 0 adapters · 0 merges
  • In apps6Hugging Face Spaces

Top Spaces running it: Eloiselatourdepise/qwen-rlcd-demo ♥ 1

The story so far

Weights, every dated article, and repo commits on one rail.

Most coverage came in the first two weeks

Teal is the weights (2026-09-16), amber triangles are 10 dated articles (each links to it), grey ticks are repo commits.

OctNovDecJanFebMarAprMayJunJulAugSepOctNovDecJanFebMarAprMayJunJulAugSepOct2026-09-16: 4 repo commitsQwen2.5-VL-32B: Smarter and Lighter · 2025-03-24Qwen2.5-Coder-32B is an LLM that can code well that runs on my Mac · 2024-11-13Qwen2.5: A Party of Foundation Models · 2024-09-18Qwen2.5-Max: Exploring the intelligence of large-scale MoE model · 2025-01-28Qwen2.5-Coder-32B with coding abilities matching those of GPT-4o · 2024-11-09Qwen2.5-Coder Beats GPT-4o and Claude 3.5 Sonnet in Coding · 2024-11-11Alibaba Qwen2.5-Omni-7B: Open Source End-to-End Multimodal AI Model · 2025-03-27Qwen2.5-VL · 2025-01-27Qwen2.5-VL-32B: Smarter and Lighter · 2025-03-25Qwen2.5-Max: The Next Leap in AI Innovation · 2025-01-29Weights 2026-09-16 · 2026-09-16Weights 2026-09-16now
OctNovDecJanFebMarAprMayJunJulAugSepOctNovDecJanFebMarAprMayJunJulAugSepOct2026-09-16: 4 repo commitsQwen2.5-VL-32B: Smarter and Lighter · 2025-03-24Qwen2.5-Coder-32B is an LLM that can code well that runs on my Mac · 2024-11-13Qwen2.5: A Party of Foundation Models · 2024-09-18Qwen2.5-Max: Exploring the intelligence of large-scale MoE model · 2025-01-28Qwen2.5-Coder-32B with coding abilities matching those of GPT-4o · 2024-11-09Qwen2.5-Coder Beats GPT-4o and Claude 3.5 Sonnet in Coding · 2024-11-11Alibaba Qwen2.5-Omni-7B: Open Source End-to-End Multimodal AI Model · 2025-03-27Qwen2.5-VL · 2025-01-27Qwen2.5-VL-32B: Smarter and Lighter · 2025-03-25Qwen2.5-Max: The Next Leap in AI Innovation · 2025-01-29Weights 2026-09-16 · 2026-09-16Weights 2026-09-16now

Reference

Everything else, folded: the full tables, every source.

Benchmarks

4 rows

Benchmarks 4 rows

measured third-party leaderboard rowauthor-reported number from the model card or eval files (4)No confidence intervals unless the source gave them; settings differ between rows, so compare within a benchmark, not across. Date = the benchmark's own date when it gives one, else when we read it.
BenchmarkValueSetting / notesReported byDate
author-reported Code Security Audit68 ms other4 fieldsmodel author2026-10-07
author-reported Enterprise Support Triage270 ms other28 fieldsmodel author2026-10-07
author-reported Fintech Fraud Routing75 ms other4 fieldsmodel author2026-10-07
author-reported High-Cardinality Tariff89 ms other1 field (255 choices)model author2026-10-07

What the card says

1 cited sentence

Summary generated, cited

The model is designed for structured information extraction, decision routing, and categorical classification using the MLX library.[1]

  1. “structured information extraction, decision routing, and …”

Generated 2026-10-07 15:17 UTC by parasail-gemma-4-26b-a4b-it; each numeral points at the fact it rests on. Not a verdict.

Use-case evidence

2 use cases

What it is good at 2 use cases

Labels are evidence, not verdicts: measured a third-party leaderboard row for this model; author-reported a number the model author published in the card or eval files; structural a deterministic signal from the name, config or chat template, or a decision-model estimate. Each chip links to its source. All use cases.

Facts

12 fields

Listed in Decision models · LLMs

Facts

Pipeline tagtext-generation
Librarymlx
MoEno
Languagesen
LicenseApache 2.0 (apache-2.0)
Hub snapshot2026-10-25

Decision facts

catalog

Decision-model catalog curated

How it was madeinference technique on stock weights (prompt + option-logit readout or constrained decoding; no new weights)
Lineage bucketW - wrapper: inference technique on stock weights or API wrapper
Lineage evidenceboard: no own weights; stock checkpoint served through own code
Base modelQwen/Qwen2.5-1.5B
DI 0.2.1 score3.78 (independent (DI 0.2.1 board))
ECE (board)0.2052
Median latency43.8 ms (DI board, 1x RTX PRO 6000, single process)
Hosted price (input)self-hosted/M
Sourcehttps://huggingface.co/harshatheg/Qwen-2.5-1B-RLCD

All variants

lineage

Lineage

Fine-tuned from Qwen/Qwen2.5-1.5B-Instruct Family: harshatheg/Qwen-2.5 0 derivatives on the Hub

Base model: Qwen/Qwen2.5-1.5B-Instruct

Repository changelog

4 commits

Repository changelog 4 commits

DateCommitSHA
2026-09-16 Add official Hugging Face Open in Spaces badge and metadata link 2af8684
2026-09-16 Add live ZeroGPU Space reference and dual MLX/PyTorch engine support 031d1a8
2026-09-16 Upload folder using huggingface_hub 6bf9d3a
2026-09-16 initial commit b4ead1e

Articles and coverage

20 pieces

Ranked by source signals (HN, Lobsters, authority domains, official posts) and recency. Summaries are generated from a stored copy of the article text; every sentence cites a span of that text, and nothing is shown that failed the citation check.

Official3

Qwen2.5-VL-32B: Smarter and Lighter
qwenlm.github.io2025-03-24authorityHN front pageabout the Qwen 2.5 familyHN 544 pts · 293 comments606 words
Alibaba open-sourced Qwen2.5-VL-32B-Instruct under the Apache 2.0 license, optimizing the model with reinforcement learning to improve human alignment and mathematical reasoning. This 32B parameter model outperforms larger 72B predecessors on benchmarks like MM-MT-Bench and MMMU-Pro while maintaining top-tier pure text capabilities. The model operates within the paradigm of "fast thinking" to prioritize subjective experience and complex multi-step reasoning.written, reviewed, cited
  • The model achieves significant advantages in MMMU, MMMU-Pro, and MathVista benchmarks focusing on complex, multi-step reasoning.
  • Qwen2.5-VL-32B-Instruct outperforms its predecessor Qwen2-VL-72B-Instruct by a substantial margin on the MM-MT-Bench benchmark.
Qwen2.5: A Party of Foundation Models
qwenlm.github.io2024-09-18authorityHN front pageabout the Qwen 2.5 familyHN 168 pts · 38 comments1,744 words
Qwen2.5 introduces dense decoder-only models pretrained on up to 18 trillion tokens, available in sizes from 0.5B to 72B parameters. Most variants are licensed under Apache 2.0, excluding the 3B and 72B models, and support a 128K context window with 8K generation limits. Specialized versions for coding and mathematics offer enhanced performance on benchmarks like HumanEval and MATH.written, reviewed, cited
  • Qwen2.5-Coder-7B-Instruct outperforms many larger language models across a range of programming languages and tasks.
  • Qwen2.5-Math-72B-Instruct general performance surpasses both Qwen2-Math-72B-Instruct and GPT4-o.
Qwen2.5-Max: Exploring the intelligence of large-scale MoE model
qwenlm.github.io2025-01-28authorityHN front pageabout the Qwen 2.5 familyHN 118 pts · 30 comments556 words
Alibaba Cloud releases Qwen2.5-Max, a large-scale MoE model pretrained on over 20 trillion tokens and post-trained with SFT and RLHF. The model outperforms DeepSeek V3 in Arena-Hard, LiveBench, LiveCodeBench, and GPQA-Diamond benchmarks. Its API, named qwen-max-2025-01-25, is OpenAI-API compatible and available via Alibaba Cloud Model Studio.written, reviewed, cited

Analysis7

Qwen2.5-Coder-32B is an LLM that can code well that runs on my Mac
simonwillison.net2024-11-13authorityHN front pageabout the Qwen 2.5 familyHN 164 pts · 65 comments635 words
Alibaba released the Qwen2.5-Coder-32B-Instruct model, an Apache 2.0 licensed open-source LLM. The 32B model runs on a 64GB MacBook Pro M2 via Ollama or MLX, fetching a 20GB quantized file. On MLX, it achieved 10.016 tokens-per-sec generation speed with 32.685 GB peak memory usage.written, reviewed, cited
  • The model requires a workaround for an SSL bug when using the standard library instead of requests or httpx.
  • Running the Ollama version was too slow because the author did not have the tool compiled to use the Mac's GPU.
Qwen2.5-Coder-32B with coding abilities matching those of GPT-4o
github.com2024-11-09authorityabout the Qwen 2.5 familyHN 21 pts · 6 comments1,815 words
Qwen2.5-Coder Beats GPT-4o and Claude 3.5 Sonnet in Coding
qwenlm.github.io2024-11-11authorityabout the Qwen 2.5 familyHN 23 pts · 0 comments1,210 words
Alibaba Qwen2.5-Omni-7B: Open Source End-to-End Multimodal AI Model
alizila.com2025-03-27about the Qwen 2.5 familyHN 29 pts · 3 comments
Qwen2.5-VL
qwenlm.github.io2025-01-27authorityabout the Qwen 2.5 familyHN 8 pts · 0 comments
Qwen2.5-VL-32B: Smarter and Lighter
simonwillison.net2025-03-25authorityabout the Qwen 2.5 familyHN 3 pts · 2 comments
Qwen2.5-Max: The Next Leap in AI Innovation
medium.com2025-01-29about the Qwen 2.5 familyHN 17 pts · 4 comments
All coverage tracked for this model (20) — append-only, nothing is ever removed
TitleSourcePublishedFirst seenSignals
PewDiePie announces his own Qwen 2.5 9B fine-tune [video]youtube.com2026-10-012026-10-07hn
Local autonomous security agent powered by Qwen 2.5-7B on Kali Linuxgithub.com2026-05-072026-10-07hn
Show HN: Using RasPi and Qwen 2.5 VL to find out when birds visit my balconyakshay326.com2026-04-042026-10-07hn
Show HN: Qwen-2.5-32B is now the best open source OCR modelgithub.com2025-04-012026-10-07hn
Qwen2.5-Omni Technical Reporthuggingface.co2025-03-302026-10-07hn
Alibaba Qwen2.5-Omni-7B: Open Source End-to-End Multimodal AI Modelalizila.com2025-03-272026-10-07hn
Qwen2.5 Omni: See, Hear, Talk, Write, Do It Allqwenlm.github.io2025-03-262026-10-07hn
Qwen2.5-VL-32B: Smarter and Lightersimonwillison.net2025-03-252026-10-07hn
Qwen2.5-VL-32B: Smarter and Lighterqwenlm.github.io2025-03-242026-10-07hn
Qwen2.5-Max: The Next Leap in AI Innovationmedium.com2025-01-292026-10-07hn
Qwen2.5-Max: Exploring the intelligence of large-scale MoE modelqwenlm.github.io2025-01-282026-10-07hn
Qwen2.5-VLqwenlm.github.io2025-01-272026-10-07hn
Qwen2.5-coder on the events occurring during the Tiananmen Square protestsnews.ycombinator.com2024-11-202026-10-07hn
Qwen2.5-Coder-32B is an LLM that can code well that runs on my Macsimonwillison.net2024-11-132026-10-07hn
Qwen2.5-Coder Technical Reportarxiv.org2024-11-112026-10-07hn
Qwen2.5-Coder Beats GPT-4o and Claude 3.5 Sonnet in Codingqwenlm.github.io2024-11-112026-10-07hn
Qwen2.5-Coder-32B with coding abilities matching those of GPT-4ogithub.com2024-11-092026-10-07hn
Guide how to use a local LLM Qwen2.5-coder-7B as an autocompelte in VSCodeeamag.me2024-10-292026-10-07hn
Run Qwen 2.5 Family and Other LMs in GGUF Format from HF Locallygithub.com2024-09-182026-10-07hn
Qwen2.5: A Party of Foundation Modelsqwenlm.github.io2024-09-182026-10-07hn

Similar models

3 models

Similar models

  • opt-125m
    same task (text-generation)
  • gpt-j-6b
    same task (text-generation), permissive license
  • DialoGPT-medium
    same task (text-generation), permissive license

Sources

3 sources

Sources

Hub snapshot: 2026-10-25 · HF API: 2026-10-24 21:43 UTC · Providers: 2026-10-24 21:44 UTC · Articles: 2026-10-24 21:46 UTC · Summary: 2026-10-07 15:17 UTC · Last verified 2026-10-25

Follow Qwen-2.5-1B-RLCD

Watch this model

Email alerts are not open yet; the feeds above are the same data. See feeds for the item schema.