- Size
- 32.8B parameterscounted from the safetensors header on the Hub
- Licence
- custom, read it, Apache License 2.0Our reading, check the licence: A custom licence we have not read for this model: other makers use the same name for different texts. Read the model's own licence file before you use it.Licence text
- On the Hub since
- 27 Apr 2025the date the repo was created
- Context length
- 40k tokensfrom the model's config
- Kind
- Base model (no base model declared on the Hub)Qwen3ForCausalLM
- On Hugging Face
- Qwen/Qwen3-32B3.9M downloads in the last 30 days, 797 likes
Scores
Epoch's name: Qwen3-32B. Best closed model on this set: gpt-6.2-astra_max, 96.8% (OpenAI). All 341 results
Epoch's name: Qwen3-32B. Best closed model on this set: claude-fable-5_high, 100.0% (Anthropic). All 318 results
- Edit format: diff40.0%8 May 2025, aider.chat
Epoch's name: Qwen3-32B. Best closed model on this set: gpt-5-2025-08-07_high, 88.0% (OpenAI). All 72 results
- default setting0.3%via Epoch AI
Epoch's name: Qwen3-32B. Best closed model on this set: gemini-4-carbon_high, 36.0% (Google DeepMind). All 271 results
- default setting67.5%conceptualreasoning.ai
Epoch's name: Qwen3-32B. Best closed model on this set: claude-fable-5-2_unknown, 99.2% (Anthropic). All 241 results
- default setting74.2%Fiction.live leaderboard
Epoch's name: chutes/Qwen3-32B. Best closed model on this set: gpt-5-2025-08-07_medium, 97.2% (OpenAI). All 61 results
- default setting17.3%12.3 to 22.3conceptualreasoning.ai
Epoch's name: Qwen3-32B. Best closed model on this set: claude-fable-5-2_xhigh, 70.6% (Anthropic). All 203 results
Claimed by the maker Claimed
The model card has no claimed scores in a machine-readable form (the card's model-index), so none are shown. The maker may give scores in the card text: read the card.
Other views, which we link to and do not copy: look this model up on Artificial Analysis and the Arena leaderboard.
Independent scores from Epoch AI (CC-BY 4.0), fetched 19 Nov 2026: Epoch AI, ‘Capabilities & benchmarking’. Published online at epoch.ai. Retrieved from ‘https://epoch.ai/benchmarks’ [online resource]. Each row links its own source. What the badges mean.
Run it yourself
Memory needed = weights + context cache + 0.50 GB for the runtime. Context cache at 8k tokens: 2.0 GB, worked out from the model's config (Qwen/Qwen3-32B), at 16-bit precision. Choose a machine above to see a verdict for each format. Method
GGUF for llama.cpp, Ollama, LM Studio and others
Qwen/Qwen3-32B-GGUF (the maker's own conversion), 115k downloads. Sizes are the real files on the Hub, all parts added up.
| Format | Size | Memory needed | Get it |
|---|---|---|---|
| Q4_K_M | 18.4 GBfile size | 20.9 GB18.4 + 2.0 + 0.50 | Commandsllama.cpp llama-server -hf Qwen/Qwen3-32B-GGUF:Q4_K_MOllama ollama run hf.co/Qwen/Qwen3-32B-GGUF:Q4_K_MDownload hf download Qwen/Qwen3-32B-GGUF --include "Qwen3-32B-Q4_K_M.gguf" --local-dir Qwen3-32B-GGUF |
| Q5_0 | 21.1 GBfile size | 23.6 GB21.1 + 2.0 + 0.50 | Commandsllama.cpp llama-server -hf Qwen/Qwen3-32B-GGUF:Q5_0Ollama ollama run hf.co/Qwen/Qwen3-32B-GGUF:Q5_0Download hf download Qwen/Qwen3-32B-GGUF --include "Qwen3-32B-Q5_0.gguf" --local-dir Qwen3-32B-GGUF |
| Q5_K_M | 21.6 GBfile size | 24.1 GB21.6 + 2.0 + 0.50 | Commandsllama.cpp llama-server -hf Qwen/Qwen3-32B-GGUF:Q5_K_MOllama ollama run hf.co/Qwen/Qwen3-32B-GGUF:Q5_K_MDownload hf download Qwen/Qwen3-32B-GGUF --include "Qwen3-32B-Q5_K_M.gguf" --local-dir Qwen3-32B-GGUF |
| Q6_K | 25.0 GBfile size | 27.5 GB25.0 + 2.0 + 0.50 | Commandsllama.cpp llama-server -hf Qwen/Qwen3-32B-GGUF:Q6_KOllama ollama run hf.co/Qwen/Qwen3-32B-GGUF:Q6_KDownload hf download Qwen/Qwen3-32B-GGUF --include "Qwen3-32B-Q6_K.gguf" --local-dir Qwen3-32B-GGUF |
| Q8_0 | 32.4 GBfile size | 34.9 GB32.4 + 2.0 + 0.50 | Commandsllama.cpp llama-server -hf Qwen/Qwen3-32B-GGUF:Q8_0Ollama ollama run hf.co/Qwen/Qwen3-32B-GGUF:Q8_0Download hf download Qwen/Qwen3-32B-GGUF --include "Qwen3-32B-Q8_0.gguf" --local-dir Qwen3-32B-GGUF |
MLX for Apple silicon
| Format | Size | Memory needed | Repo |
|---|---|---|---|
| MLX 4bitlmstudio-community/Qwen3-32B-MLX-4bit | 17.2 GBfile size | 19.7 GB17.2 + 2.0 + 0.50 | Download page |
| MLX 8bitlmstudio-community/Qwen3-32B-MLX-8bit | 32.4 GBfile size | 34.9 GB32.4 + 2.0 + 0.50 | Download page |
Original weights size from the safetensors header, for vLLM, SGLang, Transformers
| Format | Size | Memory needed | Repo |
|---|---|---|---|
| BF16original weights | 61.0 GBfile size | 63.5 GB61.0 + 2.0 + 0.50 |
hf download Qwen/Qwen3-32B --local-dir Qwen3-32BOther conversions (23; sizes not read)
- Qwen/Qwen3-32B-AWQ AWQ, by the maker, 273k downloads
- Qwen/Qwen3-32B-FP8 FP8, by the maker, 55k downloads
- unsloth/Qwen3-32B-bnb-4bit bitsandbytes, 4bit, 72k downloads
- nvidia/Qwen3-32B-NVFP4 NVFP4, 10k downloads
- unsloth/Qwen3-32B-unsloth-bnb-4bit bitsandbytes, 4bit, 4.1k downloads
- RedHatAI/Qwen3-32B-NVFP4 NVFP4, 3.6k downloads
- INCModel/Qwen3-32B-MXFP8-FP8ATTN-CT-AutoRound Other, 3.1k downloads
- INCModel/Qwen3-32B-MXFP8-FP8KV-CT-AutoRound Other, 1.6k downloads
- RedHatAI/Qwen3-32B-quantized.w4a16 Other, w4a16, 1.6k downloads
- INCModel2/Qwen3-32B-MXFP8-FP8ATTN-CT-AutoRound Other, 1.4k downloads
- RedHatAI/Qwen3-32B-FP8-dynamic FP8, 1.0k downloads
- JunHowie/Qwen3-32B-GPTQ-Int4 GPTQ, int4, 762 downloads
- defnoteatin/PseudoBrain-Qwen3-32B-NVFP4 NVFP4, 587 downloads
- defnoteatin/PseudoBrain-Qwen3-32B-W4A16 AWQ, w4a16, 566 downloads
- RedHatAI/Qwen3-32B-NVFP4A16 NVFP4, 176 downloads
- kaitchup/Qwen3-32B-autoround-4bit-gptq GPTQ, 4bit, 168 downloads
- defnoteatin/PseudoBrain-Qwen3-32B-FP8 FP8, 101 downloads
- kaitchup/Qwen3-32B-autoround-2bit-gptq GPTQ, 2bit, 84 downloads
- turboderp/Qwen3-32B-exl3 EXL3, 79 downloads
- pytorch/Qwen3-32B-FP8 FP8, 43 downloads
- JunHowie/Qwen3-32B-GPTQ-Int8 GPTQ, int8, 39 downloads
- drawais/Qwen3-32B-NVFP4 AWQ, nvfp4, 30 downloads
- Kailune-AI/Qwen3-32B-AWQ AWQ, 24 downloads
The Hub tags 100 or more quantised repos as conversions of this model; we read the most downloaded and show the maker's own first.
Run it by API
4 providers serve this model through the Hugging Face router. Prices are US dollars per million tokens. Precision is what the provider reports; a lower precision than the original weights can change answers. Live first, then cheapest output.
| Provider | In | Out | Context | Precision | Tokens/s | First token | Uptime 24h | Tools | JSON |
|---|---|---|---|---|---|---|---|---|---|
| DeepInfra | $0.08 | $0.28 | 40k | fp8 | 23 | 0.5 s | 99.8% | yes | yes |
| Tenstorrent | $0.14 | $0.40 | 20k | fp8 | 30 | 1.2 s | 90.1% | yes | no |
| SiliconFlow | $0.14 | $0.57 | 128k | fp8 | 28 | 1.3 s | 81.9% | yes | yes |
| featherless-ai | ? | ? | ? | not given | ? | ? | ? | not given | not given |
Through the router, with a Hugging Face token (the router bills your Hugging Face account). Change the name after the colon to pick another provider.
curl https://router.huggingface.co/v1/chat/completions \
-H "Authorization: Bearer $HF_TOKEN" -H "Content-Type: application/json" \
-d '{"model": "Qwen/Qwen3-32B:deepinfra", "messages": [{"role": "user", "content": "Hello"}]}'Not built yet: one key for every open model, with BlockRun, and our own tested-quality score per provider. Until then, precision and measured speed are the best guide to a provider here.
Fine-tunes of this model
Repos that declare this model as their base, most downloaded first. Listing is not a recommendation; unreviewed ones have not been checked.
- microsoft/FrogBoss-32B-25102.5kUnreviewed
- nvidia/Qwen3-Nemotron-32B-GenRM-Principle32.8B2.2kUnreviewed
- nvidia/Qwen3-Nemotron-32B-RLBFF32.8B2.1kUnreviewed
- aisingapore/Qwen-SEA-LION-v4-32B-IT32.8B1.9kUnreviewed
- inclusionAI/Qwen3-32B-AWorld32.8B1.9kUnreviewed
Planned for this page Not built yet
- Try it in private chat
- A link to Solana AI's confidential chat for the models it serves. This model is not listed as served there.
- Fine-tune it
- Fine-tuning a few base models from this page and publishing the result here.
- Creators and earnings
- Pages for people who publish fine-tunes and adapters of this model, and what they earn.
- Fund and bounties
- Bounties for new open models, paid from a public fund. How the fund will work.
- Contribute compute
- How to give compute to open model work, with Akash.
Model data from the Hugging Face Hub 19 Nov 2026, 13:43 UTC. The repo. Found a mistake on this page? Tell us.