Skip to content
Weights Not Walls
Menu

deepseek-ai

DeepSeek-V3.2-Exp

Reviewed
Size
685B parameterscounted from the safetensors header on the Hub
Licence
unknown, MIT LicenseOur reading, check the licence: You can use, change and redistribute the model for any purpose, including commercially. Keep the copyright notice and the licence text with copies. You do not have to share your changes.
On the Hub since
29 Sep 2025the date the repo was created
Context length
160k tokensfrom the model's config
Fine-tune
of deepseek-ai/DeepSeek-V3.2-Exp-BaseDeepseekV32ForCausalLM
On Hugging Face
deepseek-ai/DeepSeek-V3.2-Exp56k downloads in the last 30 days, 1.0k likes

Scores

11 independent results on 9 eval sets

Epoch's name: DeepSeek-V3.2-Exp. Best closed model on this set: gpt-5-2025-08-07_high, 88.0% (OpenAI). All 72 results

CL-benchReported

Epoch's name: DeepSeek-V3.2-Exp. Best closed model on this set: gpt-5.4-2026-03-05_xhigh, 27.9% (OpenAI). All 23 results

Epoch's name: DeepSeek-V3.2-Exp. Best closed model on this set: gpt-5.5_high, 22.2% (OpenAI). All 17 results

CritPtReported

Epoch's name: DeepSeek-V3.2-Exp. Best closed model on this set: gemini-4-carbon_high, 36.0% (Google DeepMind). All 271 results

DTBenchReported

Epoch's name: DeepSeek-V3.2-Exp. Best closed model on this set: claude-fable-5-2_unknown, 99.2% (Anthropic). All 241 results

Epoch's name: DeepSeek-V3.2-Exp. Best closed model on this set: gpt-5-2025-08-07_medium, 97.2% (OpenAI). All 61 results

LMCAReported

Epoch's name: DeepSeek-V3.2-Exp. Best closed model on this set: claude-fable-5-2_xhigh, 70.6% (Anthropic). All 203 results

Epoch's name: DeepSeek-V3.2-Exp. Best closed model on this set: gemini-2.5-flash-preview-09-2025, 41.1% (Google DeepMind). All 16 results

WeirdMLReported

Epoch's name: DeepSeek-V3.2-Exp. Best closed model on this set: gpt-6-astra_promax, 93.6% (OpenAI). All 173 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: 0.54 GB, worked out from the model's config (deepseek-ai/DeepSeek-V3.2-Exp), at 16-bit precision. Choose a machine above to see a verdict for each format. Method

Estimated sizes no measured GGUF files for this model

From the parameter count and the bits per weight of each level (the median measured over the GGUF files in our snapshot, or the format's block size). Real files differ by a few per cent.

FormatSizeMemory neededRepo
Q4_K_M4.88 bits per weight, measured median389.3 GBestimated390.3 GB389.3 + 0.54 + 0.50
Q5_K_M5.71 bits per weight, measured median455.6 GBestimated456.6 GB455.6 + 0.54 + 0.50
Q6_K6.57 bits per weight, measured median524.1 GBestimated525.2 GB524.1 + 0.54 + 0.50
Q8_08.50 bits per weight, measured median678.5 GBestimated679.6 GB678.5 + 0.54 + 0.50

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

FormatSizeMemory neededRepo
F8_E4M3original weights642.1 GBfile size643.2 GB642.1 + 0.54 + 0.50
hf download deepseek-ai/DeepSeek-V3.2-Exp --local-dir DeepSeek-V3.2-Exp
Other conversions (3; sizes not read)

The Hub tags 3 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

1 provider serves 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$0.27$0.41160kfp8162.2 s99.2%yesyes

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": "deepseek-ai/DeepSeek-V3.2-Exp: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.