Skip to content
Weights Not Walls
Menu

Qwen

Qwen3.5-122B-A10B

Reviewedaccepts images
Size
125B parameters, 10.0B active (from the repo name)counted 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
24 Feb 2026the date the repo was created
Context length
256k tokensfrom the model's config
Kind
Base model (no base model declared on the Hub)Qwen3_5MoeForConditionalGeneration
On Hugging Face
Qwen/Qwen3.5-122B-A10B738k downloads in the last 30 days, 658 likes

Scores

4 independent results on 4 eval sets
Mystery Game PuzzlesRe-run by Epoch AI
  • effort none17.0%±3.827 Aug 2026, log

Epoch's name: qwen3.5-122b-a10b. Best closed model on this set: gpt-6.2-astra_max, 95.0% (OpenAI). All 148 results

CritPtReported

Epoch's name: qwen3.5-122b-a10b. Best closed model on this set: gemini-4-carbon_high, 36.0% (Google DeepMind). All 271 results

DTBenchReported

Epoch's name: qwen3.5-122b-a10b. Best closed model on this set: claude-fable-5-2_unknown, 99.2% (Anthropic). All 241 results

LMCAReported

Epoch's name: qwen3.5-122b-a10b. 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

file sizes read from the Hub 19 Nov 2026

Memory needed = weights + context cache + 0.50 GB for the runtime. Context cache at 8k tokens: 0.19 GB, worked out from the model's config (Qwen/Qwen3.5-122B-A10B), 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

unsloth/Qwen3.5-122B-A10B-GGUF (converted by unsloth), 97k downloads. Sizes are the real files on the Hub, all parts added up. For image input you also need the projector file mmproj-F16.gguf (0.85 GB), not counted below.

FormatSizeMemory neededGet it
UD-IQ3_XXS41.7 GBfile size42.4 GB41.7 + 0.19 + 0.50
Commands
llama.cpp
llama-server -hf unsloth/Qwen3.5-122B-A10B-GGUF:UD-IQ3_XXS
Ollama
ollama run hf.co/unsloth/Qwen3.5-122B-A10B-GGUF:UD-IQ3_XXS
Download
hf download unsloth/Qwen3.5-122B-A10B-GGUF --include "Qwen3.5-122B-A10B-UD-IQ3_XXS.gguf" --local-dir Qwen3.5-122B-A10B-GGUF
UD-IQ3_S43.4 GBfile size44.0 GB43.4 + 0.19 + 0.50
Commands
llama.cpp
llama-server -hf unsloth/Qwen3.5-122B-A10B-GGUF:UD-IQ3_S
Ollama
ollama run hf.co/unsloth/Qwen3.5-122B-A10B-GGUF:UD-IQ3_S
Download
hf download unsloth/Qwen3.5-122B-A10B-GGUF --include "Qwen3.5-122B-A10B-UD-IQ3_S.gguf" --local-dir Qwen3.5-122B-A10B-GGUF
Q3_K_S48.9 GBfile size, 3 parts49.6 GB48.9 + 0.19 + 0.50
Commands
llama.cpp
llama-server -hf unsloth/Qwen3.5-122B-A10B-GGUF:Q3_K_S
Download
hf download unsloth/Qwen3.5-122B-A10B-GGUF --include "Q3_K_S/Qwen3.5-122B-A10B-Q3_K_S-*-of-00003.gguf" --local-dir Qwen3.5-122B-A10B-GGUF
Q3_K_M52.5 GBfile size, 3 parts53.2 GB52.5 + 0.19 + 0.50
Commands
llama.cpp
llama-server -hf unsloth/Qwen3.5-122B-A10B-GGUF:Q3_K_M
Download
hf download unsloth/Qwen3.5-122B-A10B-GGUF --include "Q3_K_M/Qwen3.5-122B-A10B-Q3_K_M-*-of-00003.gguf" --local-dir Qwen3.5-122B-A10B-GGUF
UD-Q3_K_XL53.1 GBfile size, 3 parts53.7 GB53.1 + 0.19 + 0.50
Commands
llama.cpp
llama-server -hf unsloth/Qwen3.5-122B-A10B-GGUF:UD-Q3_K_XL
Download
hf download unsloth/Qwen3.5-122B-A10B-GGUF --include "UD-Q3_K_XL/Qwen3.5-122B-A10B-UD-Q3_K_XL-*-of-00003.gguf" --local-dir Qwen3.5-122B-A10B-GGUF
UD-IQ4_XS56.1 GBfile size, 3 parts56.8 GB56.1 + 0.19 + 0.50
Commands
llama.cpp
llama-server -hf unsloth/Qwen3.5-122B-A10B-GGUF:UD-IQ4_XS
Download
hf download unsloth/Qwen3.5-122B-A10B-GGUF --include "UD-IQ4_XS/Qwen3.5-122B-A10B-UD-IQ4_XS-*-of-00003.gguf" --local-dir Qwen3.5-122B-A10B-GGUF
UD-IQ4_NL57.2 GBfile size, 3 parts57.9 GB57.2 + 0.19 + 0.50
Commands
llama.cpp
llama-server -hf unsloth/Qwen3.5-122B-A10B-GGUF:UD-IQ4_NL
Download
hf download unsloth/Qwen3.5-122B-A10B-GGUF --include "UD-IQ4_NL/Qwen3.5-122B-A10B-UD-IQ4_NL-*-of-00003.gguf" --local-dir Qwen3.5-122B-A10B-GGUF
Q4_K_S66.8 GBfile size, 3 parts67.5 GB66.8 + 0.19 + 0.50
Commands
llama.cpp
llama-server -hf unsloth/Qwen3.5-122B-A10B-GGUF:Q4_K_S
Download
hf download unsloth/Qwen3.5-122B-A10B-GGUF --include "Q4_K_S/Qwen3.5-122B-A10B-Q4_K_S-*-of-00003.gguf" --local-dir Qwen3.5-122B-A10B-GGUF
MXFP4_MOE69.5 GBfile size, 3 parts70.2 GB69.5 + 0.19 + 0.50
Commands
llama.cpp
llama-server -hf unsloth/Qwen3.5-122B-A10B-GGUF:MXFP4_MOE
Download
hf download unsloth/Qwen3.5-122B-A10B-GGUF --include "MXFP4_MOE/Qwen3.5-122B-A10B-MXFP4_MOE-*-of-00003.gguf" --local-dir Qwen3.5-122B-A10B-GGUF
Q4_K_M71.3 GBfile size, 3 parts72.0 GB71.3 + 0.19 + 0.50
Commands
llama.cpp
llama-server -hf unsloth/Qwen3.5-122B-A10B-GGUF:Q4_K_M
Download
hf download unsloth/Qwen3.5-122B-A10B-GGUF --include "Q4_K_M/Qwen3.5-122B-A10B-Q4_K_M-*-of-00003.gguf" --local-dir Qwen3.5-122B-A10B-GGUF
UD-Q4_K_XL71.7 GBfile size, 3 parts72.4 GB71.7 + 0.19 + 0.50
Commands
llama.cpp
llama-server -hf unsloth/Qwen3.5-122B-A10B-GGUF:UD-Q4_K_XL
Download
hf download unsloth/Qwen3.5-122B-A10B-GGUF --include "UD-Q4_K_XL/Qwen3.5-122B-A10B-UD-Q4_K_XL-*-of-00003.gguf" --local-dir Qwen3.5-122B-A10B-GGUF
Q5_K_S80.5 GBfile size, 3 parts81.1 GB80.5 + 0.19 + 0.50
Commands
llama.cpp
llama-server -hf unsloth/Qwen3.5-122B-A10B-GGUF:Q5_K_S
Download
hf download unsloth/Qwen3.5-122B-A10B-GGUF --include "Q5_K_S/Qwen3.5-122B-A10B-Q5_K_S-*-of-00003.gguf" --local-dir Qwen3.5-122B-A10B-GGUF
Q5_K_M85.2 GBfile size, 3 parts85.9 GB85.2 + 0.19 + 0.50
Commands
llama.cpp
llama-server -hf unsloth/Qwen3.5-122B-A10B-GGUF:Q5_K_M
Download
hf download unsloth/Qwen3.5-122B-A10B-GGUF --include "Q5_K_M/Qwen3.5-122B-A10B-Q5_K_M-*-of-00003.gguf" --local-dir Qwen3.5-122B-A10B-GGUF
UD-Q5_K_XL85.6 GBfile size, 3 parts86.3 GB85.6 + 0.19 + 0.50
Commands
llama.cpp
llama-server -hf unsloth/Qwen3.5-122B-A10B-GGUF:UD-Q5_K_XL
Download
hf download unsloth/Qwen3.5-122B-A10B-GGUF --include "UD-Q5_K_XL/Qwen3.5-122B-A10B-UD-Q5_K_XL-*-of-00003.gguf" --local-dir Qwen3.5-122B-A10B-GGUF
Q6_K94.1 GBfile size, 4 parts94.8 GB94.1 + 0.19 + 0.50
Commands
llama.cpp
llama-server -hf unsloth/Qwen3.5-122B-A10B-GGUF:Q6_K
Download
hf download unsloth/Qwen3.5-122B-A10B-GGUF --include "Q6_K/Qwen3.5-122B-A10B-Q6_K-*-of-00004.gguf" --local-dir Qwen3.5-122B-A10B-GGUF
UD-Q6_K_XL104.7 GBfile size, 4 parts105.4 GB104.7 + 0.19 + 0.50
Commands
llama.cpp
llama-server -hf unsloth/Qwen3.5-122B-A10B-GGUF:UD-Q6_K_XL
Download
hf download unsloth/Qwen3.5-122B-A10B-GGUF --include "UD-Q6_K_XL/Qwen3.5-122B-A10B-UD-Q6_K_XL-*-of-00004.gguf" --local-dir Qwen3.5-122B-A10B-GGUF
Q8_0121.0 GBfile size, 4 parts121.6 GB121.0 + 0.19 + 0.50
Commands
llama.cpp
llama-server -hf unsloth/Qwen3.5-122B-A10B-GGUF:Q8_0
Download
hf download unsloth/Qwen3.5-122B-A10B-GGUF --include "Q8_0/Qwen3.5-122B-A10B-Q8_0-*-of-00004.gguf" --local-dir Qwen3.5-122B-A10B-GGUF
UD-Q8_K_XL159.1 GBfile size, 5 parts159.8 GB159.1 + 0.19 + 0.50
Commands
llama.cpp
llama-server -hf unsloth/Qwen3.5-122B-A10B-GGUF:UD-Q8_K_XL
Download
hf download unsloth/Qwen3.5-122B-A10B-GGUF --include "UD-Q8_K_XL/Qwen3.5-122B-A10B-UD-Q8_K_XL-*-of-00005.gguf" --local-dir Qwen3.5-122B-A10B-GGUF
BF16227.5 GBfile size, 5 parts228.2 GB227.5 + 0.19 + 0.50
Commands
llama.cpp
llama-server -hf unsloth/Qwen3.5-122B-A10B-GGUF:BF16
Download
hf download unsloth/Qwen3.5-122B-A10B-GGUF --include "BF16/Qwen3.5-122B-A10B-BF16-*-of-00005.gguf" --local-dir Qwen3.5-122B-A10B-GGUF

MLX for Apple silicon

FormatSizeMemory neededRepo
MLXbaa-ai/Qwen3.5-122B-A10B-RAM-48GB-MLX43.9 GBfile size44.6 GB43.9 + 0.19 + 0.50Download page
MLXmlx-community/Qwen3.5-122B-A10B-mxfp8118.1 GBfile size118.8 GB118.1 + 0.19 + 0.50Download page

Original weights size from the safetensors header, for vLLM, SGLang, Transformers

FormatSizeMemory neededRepo
BF16original weights233.0 GBfile size233.7 GB233.0 + 0.19 + 0.50
hf download Qwen/Qwen3.5-122B-A10B --local-dir Qwen3.5-122B-A10B
Other conversions (17; sizes not read)

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

Hugging Face Inference Providers, fetched 19 Nov 2026, 13:58 UTC

3 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.

ProviderInOutContextPrecisionTokens/sFirst tokenUptime 24hToolsJSON
SiliconFlow$0.26$2.08256kfp8471.9 s99.9%noyes
deepinfra$0.29$2.40256knot given250.4 s?yesyes
Novita$0.40$3.20256kbf16701.8 s100.0%yesno

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/Qwen3.5-122B-A10B:siliconflow", "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.

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.