Skip to content
Weights Not Walls
Menu

deepseek-ai

DeepSeek-V2.5

Reviewed
Size
236B parameterscounted from the safetensors header on the Hub
Licence
custom, read it, DeepSeek License AgreementOur 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
5 Sep 2024the date the repo was created
Context length
160k tokensfrom the model's config
Kind
Base model (no base model declared on the Hub)DeepseekV2ForCausalLM
On Hugging Face
deepseek-ai/DeepSeek-V2.54.0k downloads in the last 30 days, 750 likes

Scores

1 independent result on 1 eval set

Epoch's name: DeepSeek-V2.5. Best closed model on this set: gpt-5-2025-08-07_high, 88.0% (OpenAI). All 72 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: 0.53 GB, worked out from the model's config (deepseek-ai/DeepSeek-V2.5), 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

legraphista/DeepSeek-V2.5-IMat-GGUF (converted by legraphista), 4.0k downloads. Sizes are the real files on the Hub, all parts added up.

FormatSizeMemory neededGet it
IQ3_XXS84.6 GBfile size, 4 parts85.6 GB84.6 + 0.53 + 0.50
Commands
llama.cpp
llama-server -hf legraphista/DeepSeek-V2.5-IMat-GGUF:IQ3_XXS
Download
hf download legraphista/DeepSeek-V2.5-IMat-GGUF --include "DeepSeek-V2.5.IQ3_XXS/DeepSeek-V2.5.IQ3_XXS-*-of-00004.gguf" --local-dir DeepSeek-V2.5-IMat-GGUF
IQ3_XS89.7 GBfile size, 5 parts90.7 GB89.7 + 0.53 + 0.50
Commands
llama.cpp
llama-server -hf legraphista/DeepSeek-V2.5-IMat-GGUF:IQ3_XS
Download
hf download legraphista/DeepSeek-V2.5-IMat-GGUF --include "DeepSeek-V2.5.IQ3_XS/DeepSeek-V2.5.IQ3_XS-*-of-00005.gguf" --local-dir DeepSeek-V2.5-IMat-GGUF
IQ3_S94.7 GBfile size, 5 parts95.7 GB94.7 + 0.53 + 0.50
Commands
llama.cpp
llama-server -hf legraphista/DeepSeek-V2.5-IMat-GGUF:IQ3_S
Download
hf download legraphista/DeepSeek-V2.5-IMat-GGUF --include "DeepSeek-V2.5.IQ3_S/DeepSeek-V2.5.IQ3_S-*-of-00005.gguf" --local-dir DeepSeek-V2.5-IMat-GGUF
Q3_K_S94.7 GBfile size, 5 parts95.7 GB94.7 + 0.53 + 0.50
Commands
llama.cpp
llama-server -hf legraphista/DeepSeek-V2.5-IMat-GGUF:Q3_K_S
Download
hf download legraphista/DeepSeek-V2.5-IMat-GGUF --include "DeepSeek-V2.5.Q3_K_S/DeepSeek-V2.5.Q3_K_S-*-of-00005.gguf" --local-dir DeepSeek-V2.5-IMat-GGUF
IQ3_M96.3 GBfile size, 5 parts97.3 GB96.3 + 0.53 + 0.50
Commands
llama.cpp
llama-server -hf legraphista/DeepSeek-V2.5-IMat-GGUF:IQ3_M
Download
hf download legraphista/DeepSeek-V2.5-IMat-GGUF --include "DeepSeek-V2.5.IQ3_M/DeepSeek-V2.5.IQ3_M-*-of-00005.gguf" --local-dir DeepSeek-V2.5-IMat-GGUF
Q3_K104.9 GBfile size, 5 parts106.0 GB104.9 + 0.53 + 0.50
Commands
llama.cpp
llama-server -hf legraphista/DeepSeek-V2.5-IMat-GGUF:Q3_K
Download
hf download legraphista/DeepSeek-V2.5-IMat-GGUF --include "DeepSeek-V2.5.Q3_K/DeepSeek-V2.5.Q3_K-*-of-00005.gguf" --local-dir DeepSeek-V2.5-IMat-GGUF
Q3_K_L114.0 GBfile size, 6 parts115.0 GB114.0 + 0.53 + 0.50
Commands
llama.cpp
llama-server -hf legraphista/DeepSeek-V2.5-IMat-GGUF:Q3_K_L
Download
hf download legraphista/DeepSeek-V2.5-IMat-GGUF --include "DeepSeek-V2.5.Q3_K_L/DeepSeek-V2.5.Q3_K_L-*-of-00006.gguf" --local-dir DeepSeek-V2.5-IMat-GGUF
IQ4_XS116.9 GBfile size, 6 parts118.0 GB116.9 + 0.53 + 0.50
Commands
llama.cpp
llama-server -hf legraphista/DeepSeek-V2.5-IMat-GGUF:IQ4_XS
Download
hf download legraphista/DeepSeek-V2.5-IMat-GGUF --include "DeepSeek-V2.5.IQ4_XS/DeepSeek-V2.5.IQ4_XS-*-of-00006.gguf" --local-dir DeepSeek-V2.5-IMat-GGUF
IQ4_NL123.8 GBfile size, 6 parts124.8 GB123.8 + 0.53 + 0.50
Commands
llama.cpp
llama-server -hf legraphista/DeepSeek-V2.5-IMat-GGUF:IQ4_NL
Download
hf download legraphista/DeepSeek-V2.5-IMat-GGUF --include "DeepSeek-V2.5.IQ4_NL/DeepSeek-V2.5.IQ4_NL-*-of-00006.gguf" --local-dir DeepSeek-V2.5-IMat-GGUF
Q4_K_S124.7 GBfile size, 6 parts125.7 GB124.7 + 0.53 + 0.50
Commands
llama.cpp
llama-server -hf legraphista/DeepSeek-V2.5-IMat-GGUF:Q4_K_S
Download
hf download legraphista/DeepSeek-V2.5-IMat-GGUF --include "DeepSeek-V2.5.Q4_K_S/DeepSeek-V2.5.Q4_K_S-*-of-00006.gguf" --local-dir DeepSeek-V2.5-IMat-GGUF
Q4_K132.7 GBfile size, 7 parts133.7 GB132.7 + 0.53 + 0.50
Commands
llama.cpp
llama-server -hf legraphista/DeepSeek-V2.5-IMat-GGUF:Q4_K
Download
hf download legraphista/DeepSeek-V2.5-IMat-GGUF --include "DeepSeek-V2.5.Q4_K/DeepSeek-V2.5.Q4_K-*-of-00007.gguf" --local-dir DeepSeek-V2.5-IMat-GGUF
Q5_K_S151.2 GBfile size, 7 parts152.2 GB151.2 + 0.53 + 0.50
Commands
llama.cpp
llama-server -hf legraphista/DeepSeek-V2.5-IMat-GGUF:Q5_K_S
Download
hf download legraphista/DeepSeek-V2.5-IMat-GGUF --include "DeepSeek-V2.5.Q5_K_S/DeepSeek-V2.5.Q5_K_S-*-of-00007.gguf" --local-dir DeepSeek-V2.5-IMat-GGUF
Q5_K155.7 GBfile size, 8 parts156.8 GB155.7 + 0.53 + 0.50
Commands
llama.cpp
llama-server -hf legraphista/DeepSeek-V2.5-IMat-GGUF:Q5_K
Download
hf download legraphista/DeepSeek-V2.5-IMat-GGUF --include "DeepSeek-V2.5.Q5_K/DeepSeek-V2.5.Q5_K-*-of-00008.gguf" --local-dir DeepSeek-V2.5-IMat-GGUF
Q6_K180.2 GBfile size, 9 parts181.3 GB180.2 + 0.53 + 0.50
Commands
llama.cpp
llama-server -hf legraphista/DeepSeek-V2.5-IMat-GGUF:Q6_K
Download
hf download legraphista/DeepSeek-V2.5-IMat-GGUF --include "DeepSeek-V2.5.Q6_K/DeepSeek-V2.5.Q6_K-*-of-00009.gguf" --local-dir DeepSeek-V2.5-IMat-GGUF
Q8_0233.4 GBfile size, 11 parts234.4 GB233.4 + 0.53 + 0.50
Commands
llama.cpp
llama-server -hf legraphista/DeepSeek-V2.5-IMat-GGUF:Q8_0
Download
hf download legraphista/DeepSeek-V2.5-IMat-GGUF --include "DeepSeek-V2.5.Q8_0/DeepSeek-V2.5.Q8_0-*-of-00011.gguf" --local-dir DeepSeek-V2.5-IMat-GGUF

MLX for Apple silicon

FormatSizeMemory neededRepo
MLXmlx-community/DeepSeek-V2.5-MLX-AQ4_1_64123.6 GBfile size124.6 GB123.6 + 0.53 + 0.50Download page

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

FormatSizeMemory neededRepo
BF16original weights439.1 GBfile size440.1 GB439.1 + 0.53 + 0.50
hf download deepseek-ai/DeepSeek-V2.5 --local-dir DeepSeek-V2.5

The Hub tags 8 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

No provider on Hugging Face Inference Providers serves this model (as of 19 Nov 2026, 13:58 UTC). Other hosts may; we only list what the router reports.

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, 13:56 UTC. The repo. Found a mistake on this page? Tell us.