- Size
- 1.7T 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
- 13 Aug 2026the date the repo was created
- Context length
- 1M tokensfrom the model's config
- Kind
- Base model (no base model declared on the Hub)DeepseekV4ForCausalLM
- On Hugging Face
- deepseek-ai/DeepSeek-V4-Pro-081346k downloads in the last 30 days, 920 likes
Scores
- effort max46.0%±5.013 Oct 2026, log
Epoch's name: deepseek-v4-pro-0813. Best closed model on this set: claude-fable-5-2_max, 79.0% (Anthropic). All 166 results
- effort max26.8%±7.019 Aug 2026, log
Epoch's name: deepseek-v4-pro-0813. Best closed model on this set: claude-fable-5-2_max, 100.0% (Anthropic). All 83 results
- effort max64.6%±2.819 Aug 2026, log
Epoch's name: deepseek-v4-pro-0813. Best closed model on this set: claude-fable-5-2_max, 95.1% (Anthropic). All 127 results
- effort max91.7%±1.518 Aug 2026, log
Epoch's name: deepseek-v4-pro-0813. Best closed model on this set: gpt-6.2-astra_max, 96.8% (OpenAI). All 341 results
- effort max33.3%±10.020 Oct 2026, log
Epoch's name: deepseek-v4-pro-0813. Best closed model on this set: claude-sonnet-5-5_max, 97.8% (Anthropic). All 47 results
- effort max43.0%±5.019 Aug 2026, Epoch AI
Epoch's name: deepseek-v4-pro-0813. Best closed model on this set: gpt-6.2-astra_max, 95.0% (OpenAI). All 148 results
- effort max98.6%±1.018 Aug 2026, log
Epoch's name: deepseek-v4-pro-0813. Best closed model on this set: claude-fable-5_high, 100.0% (Anthropic). All 318 results
- effort max52.9%±1.627 Aug 2026, log
Epoch's name: deepseek-v4-pro-0813. Best closed model on this set: claude-fable-5-2_max, 94.0% (Anthropic). All 105 results
- effort unknown47.3%via Epoch AI (error bar left out)
Epoch's name: deepseek-v4-pro-0813. Best closed model on this set: gemini-4-carbon_unknown, 84.2% (Google DeepMind). All 66 results
- effort low90.5%arcprize.org
- effort max90.0%arcprize.org
- effort high87.2%arcprize.org
- effort none13.0%arcprize.org
Epoch's name: deepseek-v4-pro-0813. Best closed model on this set: claude-fable-5_max, 98.5% (Anthropic). All 318 results
- effort max61.3%via Epoch AI
- effort high59.7%via Epoch AI
- effort low56.3%via Epoch AI
- effort none0.8%via Epoch AI
Epoch's name: deepseek-v4-pro-0813. Best closed model on this set: gpt-6-astra_max, 95.0% (OpenAI). All 299 results
- effort high18.6%via Epoch AI
- effort max18.0%via Epoch AI
- effort none0.0%via Epoch AI
Epoch's name: deepseek-v4-pro-0813. Best closed model on this set: gemini-4-carbon_high, 36.0% (Google DeepMind). All 271 results
- effort unknown93.9%conceptualreasoning.ai
Epoch's name: deepseek-v4-pro-0813. Best closed model on this set: claude-fable-5-2_unknown, 99.2% (Anthropic). All 241 results
- effort high; Harness: devin-cli; Reasoning effort: high28.5%cognition.com
Epoch's name: deepseek-v4-pro-0813. Best closed model on this set: claude-opus-5-5_medium, 54.6% (Anthropic). All 46 results
- effort unknown45.5%41.4 to 49.5conceptualreasoning.ai
Epoch's name: deepseek-v4-pro-0813. Best closed model on this set: claude-fable-5-2_xhigh, 70.6% (Anthropic). All 203 results
- effort unknown50.0%±5.0vals.ai
Epoch's name: deepseek-v4-pro-0813. Best closed model on this set: claude-fable-5-1_max, 100.0% (Anthropic). All 85 results
- effort max66.2%±0.0htihle.github.io
Epoch's name: deepseek-v4-pro-0813. 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
Memory needed = weights + context cache + 0.50 GB for the runtime. Context cache at 8k tokens: 0.95 GB, worked out from the model's config (deepseek-ai/DeepSeek-V4-Pro-0813), 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/DeepSeek-V4-Pro-0813-GGUF (converted by unsloth), 61k downloads. Sizes are the real files on the Hub, all parts added up.
| Format | Size | Memory needed | Get it |
|---|---|---|---|
| UD-Q4_K_XL | 791.3 GBfile size, 20 parts | 792.8 GB791.3 + 0.95 + 0.50 | Commandsllama.cpp llama-server -hf unsloth/DeepSeek-V4-Pro-0813-GGUF:UD-Q4_K_XLDownload hf download unsloth/DeepSeek-V4-Pro-0813-GGUF --include "UD-Q4_K_XL/DeepSeek-V4-Pro-0813-UD-Q4_K_XL-*-of-00020.gguf" --local-dir DeepSeek-V4-Pro-0813-GGUF |
| UD-Q8_K_XL | 813.5 GBfile size, 20 parts | 814.9 GB813.5 + 0.95 + 0.50 | Commandsllama.cpp llama-server -hf unsloth/DeepSeek-V4-Pro-0813-GGUF:UD-Q8_K_XLDownload hf download unsloth/DeepSeek-V4-Pro-0813-GGUF --include "UD-Q8_K_XL/DeepSeek-V4-Pro-0813-UD-Q8_K_XL-*-of-00020.gguf" --local-dir DeepSeek-V4-Pro-0813-GGUF |
Other conversions (7; sizes not read)
- nvidia/DeepSeek-V4-Pro-0813-NVFP4 NVFP4, 606 downloads
- pokey2025/dspark-DeepSeek-V4-Pro-0813.gguf GGUF, 593 downloads
- nvidia/DeepSeek-V4-Pro-0813-nvfp4-DSpark NVFP4, 526 downloads
- unsloth/DeepSeek-V4-Pro-0813 FP8, 319 downloads
- AxionML/DeepSeek-V4-Pro-0813-NVFP4 NVFP4, 121 downloads
- bullerwins/DeepSeek-V4-Pro-0813-exl3-2.04bpw EXL3, 32 downloads
- wrldsuksgo2mars/DeepSeek-V4-Pro-0813-EXL3-K2-calibrated-v1 EXL3, 31 downloads
The Hub tags 12 quantised repos as conversions of this model; we read the most downloaded and show the maker's own first.
Run it by API
13 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 |
|---|---|---|---|---|---|---|---|---|---|
| Sail Research | $1.35 | $2.40 | 1M | fp4 | 60 | 0.9 s | 100.0% | yes | yes |
| DeepInfra | $1.30 | $2.60 | 1M | fp8 | 34 | 1.4 s | 99.9% | yes | yes |
| Novita | $0.99 | $2.97 | 1M | fp8 | 58 | 1.8 s | 100.0% | yes | yes |
| BaseTen | $1.32 | $3.96 | 1M | fp4 | 68 | 0.5 s | 99.0% | yes | no |
| CoreWeave | $1.31 | $3.96 | 1M | fp8 | 135 | 0.6 s | 98.4% | yes | yes |
| DigitalOcean | $1.32 | $3.96 | 1M | not given | 94 | 1.3 s | 100.0% | yes | yes |
| Nebius | $1.32 | $3.96 | 1M | fp8 | 43 | 0.5 s | 99.9% | yes | yes |
| Parasail | $1.32 | $3.96 | 1M | fp8 | 41 | 0.6 s | 99.1% | yes | yes |
| SiliconFlow | $1.32 | $3.96 | 1M | fp8 | 25 | 1.0 s | 100.0% | yes | no |
| Together | $1.32 | $3.96 | 1M | not given | 62 | 0.7 s | 99.9% | yes | yes |
| Wafer | $0.11 | $4.44 | 1M | mxfp4 | 65 | 0.3 s | 99.9% | yes | yes |
| Venice | $1.65 | $4.95 | 977k | not given | 22 | 1.5 s | 98.0% | yes | yes |
| Relace | $0.10 | $8.00 | 1M | fp4 | 17 | 0.9 s | 99.9% | yes | no |
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-V4-Pro-0813:sail-research", "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.