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 › Qwen3 family

Qwen3-235B-A22B-Thinking-2507

Hugging Face ↗ Follow changes

Qwen Apache 2.0 out 15 months ago

235B parameters, 22B active (8 of 128 experts per token), text in, text out, 256K tokens of context. 3 providers serve it from $0.23 per million input tokens.

Verified 2026-10-25 Rank #239 in LLMs · score 86.5 method v1 Sources 4 Downloads 30d 14.8k · #479 of 619 tracked

Can I run it?141GB at INT4Estimatefits 2×80 GBmethod
What does it cost?$0.23/1M incheapest of 2DeepInfra2026-10-25
Can I use it?Apache 2.0Permissive; patent grantlicence file2026-10-24
Do people use it?14.8kdownloads, 30 days#479 of 619 we track · below the median+2 likes, downloads −18% this week418 likesHugging Face2026-10-24

Worth reading

Picked by rule from 19 tracked pieces: the lab's own post, the most shared, a field report, the latest.

Most discussed · 490 HN pts

GPT-OSS vs. Qwen3 and a detailed look how things evolved since GPT-2

OpenAI released gpt-oss-120b and gpt-oss-20b, their first open-weight models since GPT-2 in 2019.

  • A 2025 research paper confirms Dropout results in worse downstream performance in single-epoch regimes.
magazine.sebastianraschka.com 2025-08-10 490 HN pts · 97 comments 5.2k words about the Qwen3 family family

19 best of 19 tracked

The 4 picks above, plus these 15. Top is the radar's score: discussion, source authority and the lab's own posts, fading with age.

  1. ArticleQwen3-Coder-Nextqwen.ai · 2026-02-03 · official
  2. ArticleQwen3-Max-Thinkingqwen.ai · 2026-01-26 · official
  3. ArticleQwen3-TTS family is now open sourced: Voice design, clone, and generationqwen.ai · 2026-01-22 · official
  4. ArticleQwen3: Think deeper, act fasterqwenlm.github.io · 2025-04-28 · 869 HN pts · official
  5. ArticleQwen3-Coder: Agentic coding in the worldqwenlm.github.io · 2025-07-22 · official
  6. ArticleQwen3-Nextqwen.ai · 2025-09-12 · official
  7. ArticleQwen3-VLqwen.ai · 2025-09-23 · official
  8. CodeQwen3-Omni: Native Omni AI model for text, image and videogithub.com · 2025-09-22
  9. ArticleQwen3-235B-A22B-Thinking-2507simonwillison.net · 2025-07-25 · 3 HN pts
  10. ArticleQwen 3 235B-A22B Review: The MoE Flagship Tested | BestLLMforbestllmfor.com · undated
  11. ArticleQwen3 235B A22B Thinking 2507 - Specs, Capabilities & Benchmarks | Inferbaseinferbase.ai · undated
  12. ArticleQwen: Qwen3 235B A22B Thinking 2507 - AI Model Details & ...benchable.ai · undated
  13. ArticleQwen3 235B A22B Thinking 2507: benchmarks & pricing — ModelCap #85modelcap.ai · undated
  14. ArticleQwen3 235B A22B Thinking (2507)opper.ai · undated
  15. ArticleQwen3-235B-A22B-Thinking-2507 by Qwen | The GTM Directorythegtmdirectory.com · 2025-07-25

Where to run it

Live hosted endpoints, cheapest first, USD per million tokens. Ties break on output price, then provider name.

Prices for the same weights vary 2×

2 providers with a live price, from $0.23 to $0.45 per million input tokens; the median is $0.34. One dot per provider, cheapest in teal.

$0.010$0.10$1.00$10.00DeepInfra: $0.23 in / $2.30 out per 1M · observed 2026-10-25 · https://huggingface.co/api/models/Qwen/Qwen3-235B-A22B-Thinking-2507?expand[]=inferenceProviderMappingVenice: $0.45 in / $3.50 out per 1M · observed 2026-10-24 · https://openrouter.ai/qwen/qwen3-235b-a22b-thinking-2507cheapest $0.23median $0.34
$0.010$0.10$1.00$10.00DeepInfra: $0.23 in / $2.30 out per 1M · observed 2026-10-25 · https://huggingface.co/api/models/Qwen/Qwen3-235B-A22B-Thinking-2507?expand[]=inferenceProviderMappingVenice: $0.45 in / $3.50 out per 1M · observed 2026-10-24 · https://openrouter.ai/qwen/qwen3-235b-a22b-thinking-2507cheapest $0.23 · median $0.34 (white tick)
#Cheapest 3In /1MOut /1MContextQuantSeen
1 DeepInfra $0.23 $2.30 256K — Oct 25
2 Venice $0.45 $3.50 125K fp8 Oct 24
3 Featherless — — — — Oct 25

All 3 providers →

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.

Served by 3 API providers; 46 quantized builds for local use.

  • As a service3DeepInfra, Featherless, Venice
  • On own hardware46quantized builds
  • Built on1917 fine-tunes · 2 adapters · 0 merges
  • In apps59Hugging Face Spaces

Top Spaces running it: jairwaal/image ♥ 16 · umint/openwebui ♥ 6 · mgbam/builder ♥ 5 · Jakob08/moneychatbot ♥ 3

Will it fit?

Weights only, an estimate: leave room for the KV cache. Method

Fits two 80 GB GPUs at INT4; too big for one

About 141 GB at INT4, 282 GB at FP8, 564 GB at BF16: weights only, so leave room for the KV cache. Every expert must be resident, so total parameters count.

16 GB card64 GB Mac2×80 GB8×80 GBINT4: about 141 GB of weights (estimate; KV cache and activations extra)INT4141 GBFP8: about 282 GB of weights (estimate; KV cache and activations extra)FP8282 GBBF16: about 564 GB of weights (estimate; KV cache and activations extra)BF16564 GB
16 GB card64 GB Mac2×80 GB8×80 GBINT4: about 141 GB of weights (estimate; KV cache and activations extra)INT4141 GBFP8: about 282 GB of weights (estimate; KV cache and activations extra)FP8282 GBBF16: about 564 GB of weights (estimate; KV cache and activations extra)BF16564 GB

Estimate params × bytes per param × 1.2 · how we estimate

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 (2025-07-25), amber triangles are 4 dated articles (each links to it), grey ticks are repo commits.

MayJunJulAugSepOctNovDecJanFebMarAprMayJunJulAugSepOct2025-08-17: 1 repo commit2025-08-08: 1 repo commit2025-07-31: 1 repo commit2025-07-30: 1 repo commit2025-07-29: 1 repo commit2025-07-26: 1 repo commit2025-07-25: 10 repo commitsQwen3.5 122B and 35B models offer Sonnet 4.5 performance on local computers · 2026-02-28Qwen3: Think deeper, act faster · 2025-04-28GPT-OSS vs. Qwen3 and a detailed look how things evolved since GPT-2 · 2025-08-10Qwen3-235B-A22B-Thinking-2507 · 2025-07-25Weights 2025-07-25 · 2025-07-25Weights 2025-07-25now · 3 providers
MayJunJulAugSepOctNovDecJanFebMarAprMayJunJulAugSepOct2025-08-17: 1 repo commit2025-08-08: 1 repo commit2025-07-31: 1 repo commit2025-07-30: 1 repo commit2025-07-29: 1 repo commit2025-07-26: 1 repo commit2025-07-25: 10 repo commitsQwen3.5 122B and 35B models offer Sonnet 4.5 performance on local computers · 2026-02-28Qwen3: Think deeper, act faster · 2025-04-28GPT-OSS vs. Qwen3 and a detailed look how things evolved since GPT-2 · 2025-08-10Qwen3-235B-A22B-Thinking-2507 · 2025-07-25Weights 2025-07-25 · 2025-07-25Weights 2025-07-25now · 3 providers

What people built on it

65 derivatives on the Hub. Which build to run, by runtime: the most downloaded in each class.

65 derivatives on the Hub, most of them quantized

From the Hugging Face model tree; quantized builds in teal, fine-tunes, merges and adapters in greys.

46 quantized (Hugging Face model tree)17 fine-tunes (Hugging Face model tree)2 adapters (Hugging Face model tree)46 quantized17 fine-tunes2 adapters
46 quantized (Hugging Face model tree)17 fine-tunes (Hugging Face model tree)2 adapters (Hugging Face model tree)46 quantized17 fine-tunes2 adapters
GGUF unsloth/Qwen3-235B-A22B-Thinking-2507-GGUF for llama.cpp, Ollama, LM Studio 158.2k downloads MLX lmstudio-community/Qwen3-235B-A22B-Thinking-2507-MLX-8bit for Apple silicon 12.8k downloads FP8 Qwen/Qwen3-235B-A22B-Thinking-2507-FP8 for vLLM, SGLang 534k downloads

All variants →

Reference

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

Benchmarks

25 rows

Benchmarks 25 rows

measured third-party leaderboard rowauthor-reported number from the model card or eval files (25)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 AIME 2025 · Reason through hard problems92.3—model author2026-10-07
author-reported Arena-Hard · Chat and follow instructions79.7—model author2026-10-07
author-reported BFCL · Call tools and APIs71.9—model author2026-10-07
author-reported CFEval2134—model author2026-10-07
author-reported Creative Writing · Chat and follow instructions86.1—model author2026-10-07
author-reported Creative Writing · Chat and follow instructions88.3—model author2026-10-07
author-reported GPQA Diamond · Reason through hard problems81.1—model author2026-10-07
author-reported HMMT · Reason through hard problems83.9—model author2026-10-07
author-reported Humanity's Last Exam · Reason through hard problems18.2#—model author2026-10-07
author-reported IFEval · Chat and follow instructions87.8—model author2026-10-07
author-reported INCLUDE · Work across languages81—model author2026-10-07
author-reported LiveBench · Reason through hard problems78.4—model author2026-10-07
author-reported LiveCodeBench · Write code with agents74.1—model author2026-10-07
author-reported MMLU · Reason through hard problems93.8—model author2026-10-07
author-reported MMLU-Pro · Reason through hard problems84.4—model author2026-10-07
author-reported MMLU-ProX · Work across languages81—model author2026-10-07
author-reported Multi-IF · Chat and follow instructions80.6—model author2026-10-07
author-reported OJBench32.5—model author2026-10-07
author-reported PolyMATH · Work across languages60.1—model author2026-10-07
author-reported SuperGPQA · Reason through hard problems64.9—model author2026-10-07
author-reported TAU1-Airline46—model author2026-10-07
author-reported TAU1-Retail67.8—model author2026-10-07
author-reported tau-bench · Call tools and APIs71.9—model author2026-10-07
author-reported tau-bench · Call tools and APIs58—model author2026-10-07
author-reported tau-bench · Call tools and APIs45.6—model author2026-10-07

What the card says

2 cited sentences

Summary generated, cited

This mixture-of-experts architecture utilizes 128 experts and supports a 262,144 native context length.[1][2][3] The model is released under an Apache 2.0 license and is designed for complex reasoning tasks.[4][5]

  1. is mixture of experts
  2. num experts
  3. “262,144 natively”
  4. license
  5. “We strongly recommend its use in highly complex reasoning…”

Generated 2026-10-07 15:04 UTC by parasail-gemma-4-26b-a4b-it; each numeral points at the fact it rests on. Not a verdict.

Use-case evidence

6 use cases

What it is good at 6 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

24 fields

Listed in LLMs · MoE

Facts

Pipeline tagtext-generation
Librarytransformers
ArchitecturesQwen3MoeForCausalLM
Model typeqwen3_moe
Dtypebfloat16
Params by dtypeBF16: 235,093,634,560
Size bandXXL
MoEyes
Experts128
Experts per token8
Context length262,144 (config.json max_position_embeddings)
PapersarXiv:2402.17463, arXiv:2407.02490, arXiv:2501.15383, arXiv:2404.06654, arXiv:2505.09388
LicenseApache 2.0 (apache-2.0)
Hub snapshot2026-10-25

Servability

vLLM · SGLang

Servability

vLLM: supported · Qwen3MoeForCausalLM · 7050ef881ff0835f6890a47db1859a24c65481d6SGLang: supported · Qwen3MoeForCausalLM · 773cb938fe9484a8f7e7e46fb1b1fe96412c59ed

vLLM registry (7050ef881ff0835f6890a47db1859a24c65481d6) · SGLang registry (773cb938fe9484a8f7e7e46fb1b1fe96412c59ed) · Matched by architecture class name against each project's registry on main; 'supported' means the architecture is registered, not that this checkpoint was tested.

Hardware fit

3 estimates

Hardware fit ESTIMATE

Weights-only memory for Qwen3-235B-A22B-Thinking-2507: params × bytes/param × 1.2 overhead. KV cache and activations are extra. How this is computed.

BF16 (2 bytes/param)564 GB · Multi-GPU node (<640 GB)
FP8 (1 byte/param)282 GB · Multi-GPU node (<640 GB)
INT4 (0.5 bytes/param)141 GB · Multi-GPU node (<640 GB)

Weights only (params × bytes/param × 1.2 overhead); KV cache and activations extra. MoE: all experts must be resident, so total parameters count.

All variants

40 variants

Lineage

Original model Family: Qwen/Qwen3 65 derivatives on the Hub

quantized: 46finetune: 17adapter: 2

Top variants (40 shown, ranked by downloads)

Relation
Quant
Size
40 rows
Variant Relation Quant Params Band License Author Likes Downloads
Qwen/Qwen3-235B-A22B-Thinking-2507-FP8 quantized -FP8 235B XXL apache-2.0 Qwen 97 534k
unsloth/Qwen3-235B-A22B-Thinking-2507-GGUF quantized GGUF — — apache-2.0 unsloth 84 158.2k
nvidia/Qwen3-235B-A22B-Thinking-2507-NVFP4 quantized NVFP4 120B XXL apache-2.0 nvidia 8 29.5k
QuantTrio/Qwen3-235B-A22B-Thinking-2507-AWQ quantized AWQ 235B XXL apache-2.0 QuantTrio 6 25.1k
nvidia/Qwen3-Nemotron-235B-A22B-GenRM finetune — 235B XXL apache-2.0 nvidia 33 18.8k
bartowski/Qwen_Qwen3-235B-A22B-Thinking-2507-GGUF quantized GGUF — — — bartowski 3 18k
miromind-ai/MiroThinker-v1.5-235B finetune — 235B XXL mit miromind-ai 254 3.4k
miromind-ai/MiroThinker-1.7-mini finetune — 30.5B L apache-2.0 miromind-ai 104 10.5k
miromind-ai/MiroThinker-1.7 finetune — 235B XXL apache-2.0 miromind-ai 146 7.2k
lmstudio-community/Qwen3-235B-A22B-Thinking-2507-MLX-8bit quantized MLX 235B XXL apache-2.0 lmstudio-community 1 12.8k
mlx-community/Qwen3-235B-A22B-Thinking-2507-4bit quantized -4bit 235B XXL apache-2.0 mlx-community 1 12.5k
ubergarm/Qwen3-235B-A22B-Thinking-2507-GGUF quantized GGUF — — apache-2.0 ubergarm 10 10.4k
nvidia/Qwen3-Nemotron-235B-A22B-GenRM-2603 finetune — 235B XXL apache-2.0 nvidia 30 7.1k
Mungert/MiroThinker-1.7-mini-GGUF quantized GGUF — — apache-2.0 Mungert 0 8.2k
lmstudio-community/Qwen3-235B-A22B-Thinking-2507-GGUF quantized GGUF — — — lmstudio-community 2 7.9k
mlx-community/Qwen3-235B-A22B-Thinking-2507-8bit quantized -8bit 235B XXL apache-2.0 mlx-community 4 7.5k
lmstudio-community/Qwen3-235B-A22B-Thinking-2507-MLX-4bit quantized MLX 235B XXL apache-2.0 lmstudio-community 1 6.8k
Jinx-org/Jinx-Qwen3-235B-A22B-Thinking-2507 finetune — 235B XXL apache-2.0 Jinx-org 4 4.9k
lmstudio-community/Qwen3-235B-A22B-Thinking-2507-MLX-6bit quantized MLX 235B XXL apache-2.0 lmstudio-community 1 4.9k
John1604/Qwen3-235B-A22B-Thinking-2507-gguf quantized gguf — — apache-2.0 John1604 1 3.2k
mlx-community/Qwen3-235B-A22B-Thinking-2507-3bit-DWQ quantized — 235B XXL apache-2.0 mlx-community 6 2.9k
DevQuasar/Qwen.Qwen3-235B-A22B-Thinking-2507-GGUF quantized GGUF — — — DevQuasar 0 3k
lovedheart/Qwen3-235B-A22B-Thinking-2507-GGUF-IQ1-M quantized GGUF — — apache-2.0 lovedheart 2 2.8k
NVFP4/Qwen3-235B-A22B-Thinking-2507-FP4 quantized — 118B XXL apache-2.0 NVFP4 2 2k
mlx-community/Qwen3-235B-A22B-Thinking-2507-4bit-DWQ quantized -4bit 235B XXL apache-2.0 mlx-community 4 1.7k
mradermacher/Qwen3-235B-A22B-Thinking-2507-i1-GGUF quantized GGUF — — apache-2.0 mradermacher 1 1.7k
snwy/frankenqwen3-8B-235B-dense-conversion-interleaved-untuned finetune — 13.6B L apache-2.0 snwy 19 704
QuantTrio/Qwen3-235B-A22B-Thinking-2507-GPTQ-Int4-Int8Mix quantized GPTQ 253B XXL apache-2.0 QuantTrio 4 1.3k
unsloth/Qwen3-235B-A22B-Thinking-2507 finetune — 235B XXL apache-2.0 unsloth 3 1.4k
amd/Qwen3-235B-A22B-Thinking-2507-ptpc quantized — 235B XXL apache-2.0 amd 0 1.4k
nvidia/Qwen3-235B-A22B-Thinking-2507-Eagle3 finetune — 294M S other nvidia 2 1.3k
mlx-community/Qwen3-235B-A22B-Thinking-2507-3bit quantized — 235B XXL apache-2.0 mlx-community 1 869
mlx-community/Qwen3-235B-A22B-Thinking-2507-mixed-3-5bit quantized — 235B XXL apache-2.0 mlx-community 1 781
sm54/Qwen3-235B-A22B-Thinking-2507-MXFP4_MOE quantized GGUF — — — sm54 1 768
inclusionAI/AReaL-SEA-235B-A22B finetune — 235B XXL apache-2.0 inclusionAI 5 503
anikifoss/Qwen3-235B-A22B-Thinking-2507-HQ4_K quantized GGUF — — apache-2.0 anikifoss 1 636
mo386/Qwen3-235B-A22B-Thinking-2507-mlx-8Bit quantized mlx 235B XXL apache-2.0 mo386 0 524
mlx-community/Qwen3-235B-A22B-Thinking-2507-5bit quantized — 235B XXL apache-2.0 mlx-community 0 506
sjug/Qwen3-235B-A22B-Thinking-2507-MLX-9bit quantized MLX 235B XXL apache-2.0 sjug 0 502
Wwayu/Qwen3-235B-A22B-Thinking-2507-mlx-2Bit quantized mlx 235B XXL apache-2.0 Wwayu 0 495

Repository changelog

16 commits

Repository changelog 16 commits

DateCommitSHA
2025-08-17 Update README.md 6cbffae
2025-08-08 Add 1M support (#6) ddee1c5
2025-07-31 Update LICENSE d24ea30
2025-07-30 Update README.md 37aa280
2025-07-29 Update README.md 90794e4
2025-07-26 Create LICENSE 6203b69
2025-07-25 Update README.md 4cd6884
2025-07-25 Update tokenizer_config.json bff620f
2025-07-25 Update README.md 1ff45b2
2025-07-25 Create generation_config.json 2101cf0
2025-07-25 Update config.json 37fe633
2025-07-25 Delete configuration.json b95c105
2025-07-25 Delete .mv 1ff67ec
2025-07-25 Delete .msc 4429362
2025-07-25 Add files using upload-large-folder tool 8546b01
2025-07-25 initial commit 501220d

All providers

3 providers · 3 endpoints

Where to run 3 providers · 3 endpoints

Hosted endpoints observed for Qwen3-235B-A22B-Thinking-2507, sorted by input price (USD per million tokens). Each row names its observer and date; prices change daily, so check the source before quoting. Cheapest observed input price: $0.23/M.

ProviderModel refContextQuant$/M in$/M outtok/sTTFTToolsStatusObserved
DeepInfra
hf
 Qwen/Qwen3-235B-A22B-Thinking-2507 256K — $0.23 $2.30 145.3 264 ms yes live 2026-10-25
Venice
openrouter
 qwen/qwen3-235b-a22b-thinking-2507 125K fp8 $0.45 $3.50 — — yes live 2026-10-24
Featherless
hf
 Qwen/Qwen3-235B-A22B-Thinking-2507 — — — — — — — live 2026-10-25

Articles and coverage

19 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.

Official2

Qwen3: Think deeper, act faster
qwenlm.github.io2025-04-28officialauthorityHN front pageabout the Qwen3 familyHN 869 pts · 388 comments2,139 words
Qwen3 introduces Hybrid Thinking Modes, allowing users to toggle between step-by-step reasoning and rapid responses via computational budget control. The open-weight MoE model Qwen3-235B-A22B achieves performance similar to dense models while using only 10% of the active parameters. Pre-training utilized approximately 36 trillion tokens across 119 languages, with deployment supported by SGLang and vLLM.written, reviewed, cited
  • Qwen3-8B and Qwen3-14B models do not use tied embeddings, unlike smaller variants.
  • The pre-training pipeline includes a final stage extending context length to 32K tokens.
qwen3-235b-a22b-thinking-2507 Model Info - Alibaba Cloud Model Studio - Alibaba Cloud Documentation Center
alibabacloud.comofficial

Analysis8

Qwen3.5 122B and 35B models offer Sonnet 4.5 performance on local computers
venturebeat.com2026-02-28HN front pageabout the Qwen3 familyHN 461 pts · 272 comments
GPT-OSS vs. Qwen3 and a detailed look how things evolved since GPT-2
magazine.sebastianraschka.com2025-08-10authorityHN front pageabout the Qwen3 familyHN 490 pts · 97 comments5,195 words
OpenAI released gpt-oss-120b and gpt-oss-20b, their first open-weight models since GPT-2 in 2019. The 20B model can run on a consumer GPU with up to 16 GB of RAM, and the 120B model can run on a single H100 with 80 GB of RAM or newer hardware. The models utilize MXFP4 optimization to fit onto single GPUs.written, reviewed, cited
  • A 2025 research paper confirms Dropout results in worse downstream performance in single-epoch regimes.
Qwen3-235B-A22B-Thinking-2507
simonwillison.net2025-07-25authorityHN 3 pts · 0 comments
Qwen3 235B A22B Thinking 2507 · Benchmarks, Pricing & Performance | BenchGecko
benchgecko.ai
Qwen 3 235B-A22B Review: The MoE Flagship Tested | BestLLMfor
bestllmfor.com
Qwen3 235B A22B Thinking 2507 - Specs, Capabilities & Benchmarks | Inferbase
inferbase.ai
Qwen: Qwen3 235B A22B Thinking 2507 - AI Model Details & ...
benchable.ai
Qwen3 235B A22B Thinking 2507: benchmarks & pricing — ModelCap #85
modelcap.ai
All coverage tracked for this model (19) — append-only, nothing is ever removed
TitleSourcePublishedFirst seenSignals
qwen3-235b-a22b-thinking-2507 Model Info - Alibaba Cloud Model Studio - Alibaba Cloud Documentation Centeralibabacloud.com official—2026-10-07exa
Qwen3 235B A22B Thinking 2507 · Benchmarks, Pricing & Performance | BenchGeckobenchgecko.ai—2026-10-07exa
Qwen 3 235B-A22B Review: The MoE Flagship Tested | BestLLMforbestllmfor.com—2026-10-07exa
Qwen3 235B A22B Thinking 2507 - Specs, Capabilities & Benchmarks | Inferbaseinferbase.ai—2026-10-07exa
Qwen: Qwen3 235B A22B Thinking 2507 - AI Model Details & ...benchable.ai—2026-10-07exa
Qwen3 235B A22B Thinking 2507: benchmarks & pricing — ModelCap #85modelcap.ai—2026-10-07exa
Qwen3 235B A22B Thinking (2507)opper.ai—2026-10-07exa
Qwen3.5 122B and 35B models offer Sonnet 4.5 performance on local computersventurebeat.com2026-02-282026-10-07hn
Qwen3-Coder-Nextqwen.ai official2026-02-032026-10-07hn
Qwen3-Max-Thinkingqwen.ai official2026-01-262026-10-07hn
Qwen3-TTS family is now open sourced: Voice design, clone, and generationqwen.ai official2026-01-222026-10-07hn
Qwen3-VLqwen.ai official2025-09-232026-10-07hn
Qwen3-Omni: Native Omni AI model for text, image and videogithub.com2025-09-222026-10-07hn
Qwen3-Nextqwen.ai official2025-09-122026-10-07hn
GPT-OSS vs. Qwen3 and a detailed look how things evolved since GPT-2magazine.sebastianraschka.com2025-08-102026-10-07hn
Qwen3-235B-A22B-Thinking-2507simonwillison.net2025-07-252026-10-07hn
Qwen3-235B-A22B-Thinking-2507 by Qwen | The GTM Directorythegtmdirectory.com2025-07-252026-10-07exa
Qwen3-Coder: Agentic coding in the worldqwenlm.github.io official2025-07-222026-10-07hn
Qwen3: Think deeper, act fasterqwenlm.github.io official2025-04-282026-10-07hn

Papers

5 papers

Papers 5

arXiv links found in the model card and its tags. "Peer-reviewed" is shown only with evidence: a published venue in OpenAlex (measured) or the authors' own journal reference on arXiv (reported).

Similar models

3 models

Similar models

  • MiniMax-M2.5 · 229B
    same size band (XXL), also MoE, same task (text-generation), more providers
  • MiniMax-M2.7 · 229B
    same size band (XXL), also MoE, same task (text-generation), more providers
  • MiniMax-M2 · 229B
    same size band (XXL), also MoE, same task (text-generation)

Sources

4 sources

Sources

Hub snapshot: 2026-10-25 · HF API: 2026-10-24 21:36 UTC · Providers: 2026-10-24 21:44 UTC · Articles: 2026-10-24 21:45 UTC · Summary: 2026-10-07 15:04 UTC · Last verified 2026-10-25

Follow Qwen3-235B-A22B-Thinking-2507

Watch this model

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