Skip to content
Weights Not Walls
Menu

deepseek-ai

DeepSeek-V4.1-Flash

Reviewedaccepts images
Size
763B 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
10 Sep 2026the date the repo was created
Context length
1M tokensfrom the model's config
Kind
Base model (no base model declared on the Hub)DeepseekV41ForCausalLM
On Hugging Face
deepseek-ai/DeepSeek-V4.1-Flash1.5M downloads in the last 30 days, 5.0k likes

Scores

28 independent results on 21 eval sets
Chess PuzzlesRe-run by Epoch AI
  • effort max48.0%±5.013 Oct 2026, log

Epoch's name: deepseek-v4.1-flash. Best closed model on this set: claude-fable-5-2_max, 79.0% (Anthropic). All 166 results

  • effort max26.8%±7.08 Oct 2026, Epoch AI

Epoch's name: deepseek-v4.1-flash. Best closed model on this set: claude-fable-5-2_max, 100.0% (Anthropic). All 83 results

  • effort max67.4%±2.88 Oct 2026, Epoch AI

Epoch's name: deepseek-v4.1-flash. Best closed model on this set: claude-fable-5-2_max, 95.1% (Anthropic). All 127 results

Furniture AssemblyRe-run by Epoch AI
  • effort max34.2%±6.08 Oct 2026, Epoch AI

Epoch's name: deepseek-v4.1-flash. Best closed model on this set: claude-fable-5-2_max, 90.0% (Anthropic). All 41 results

GPQA diamondRe-run by Epoch AI
  • effort max89.8%±1.79 Oct 2026, log

Epoch's name: deepseek-v4.1-flash. Best closed model on this set: gpt-6.2-astra_max, 96.8% (OpenAI). All 341 results

MirrorCodeRe-run by Epoch AI
  • effort max64.4%±11.421 Oct 2026, Epoch AI

Epoch's name: deepseek-v4.1-flash. Best closed model on this set: claude-sonnet-5-5_max, 97.8% (Anthropic). All 47 results

Mystery Game PuzzlesRe-run by Epoch AI
  • effort max43.0%±5.08 Oct 2026, Epoch AI

Epoch's name: deepseek-v4.1-flash. Best closed model on this set: gpt-6.2-astra_max, 95.0% (OpenAI). All 148 results

OTIS Mock AIME 2024-2025Re-run by Epoch AI
  • effort max98.3%±0.68 Oct 2026, log

Epoch's name: deepseek-v4.1-flash. Best closed model on this set: claude-fable-5_high, 100.0% (Anthropic). All 318 results

SimpleQA VerifiedRe-run by Epoch AI
  • effort max42.9%±1.613 Oct 2026, log

Epoch's name: deepseek-v4.1-flash. Best closed model on this set: claude-fable-5-2_max, 94.0% (Anthropic). All 105 results

APEX-AgentsReported

Epoch's name: deepseek-v4.1-flash. Best closed model on this set: gemini-4-carbon_unknown, 84.2% (Google DeepMind). All 66 results

ARC-AGIReported

Epoch's name: deepseek-v4.1-flash. Best closed model on this set: claude-fable-5_max, 98.5% (Anthropic). All 318 results

ARC-AGI-2Reported

Epoch's name: deepseek-v4.1-flash. Best closed model on this set: gpt-6-astra_max, 95.0% (OpenAI). All 299 results

CritPtReported

Epoch's name: deepseek-v4.1-flash. Best closed model on this set: gemini-4-carbon_high, 36.0% (Google DeepMind). All 271 results

DTBenchReported

Epoch's name: deepseek-v4.1-flash. Best closed model on this set: claude-fable-5-2_unknown, 99.2% (Anthropic). All 241 results

FrontierCodeReported
  • effort high; Harness: devin-cli; Reasoning effort: high43.9%cognition.com

Epoch's name: deepseek-v4.1-flash. Best closed model on this set: claude-opus-5-5_medium, 54.6% (Anthropic). All 46 results

LMCAReported

Epoch's name: deepseek-v4.1-flash. Best closed model on this set: claude-fable-5-2_xhigh, 70.6% (Anthropic). All 203 results

ProofBenchReported

Epoch's name: deepseek-v4.1-flash. Best closed model on this set: claude-fable-5-1_max, 100.0% (Anthropic). All 85 results

Epoch's name: deepseek-v4.1-flash. Best closed model on this set: claude-fable-5-2_unknown, 41.3% (Anthropic). All 20 results

SimpleBenchReported

Epoch's name: deepseek-v4.1-flash. Best closed model on this set: claude-fable-5-2_unknown, 94.9% (Anthropic). All 125 results

Epoch's name: deepseek-v4.1-flash. Best closed model on this set: claude-fable-5_high, 95.0% (Anthropic). All 29 results

WeirdML v3Reported

Epoch's name: deepseek-v4.1-flash. Best closed model on this set: claude-fable-5-2_xhigh, 65.3% (Anthropic). All 27 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: 0.63 GB, worked out from the model's config (deepseek-ai/DeepSeek-V4.1-Flash), 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

antirez/deepseek-v4.1-flash-gguf (converted by antirez), 767k downloads. Sizes are the real files on the Hub, all parts added up.

FormatSizeMemory neededGet it
Q2340.6 GBfile size341.7 GB340.6 + 0.63 + 0.50
Commands
llama.cpp
llama-server -hf antirez/deepseek-v4.1-flash-gguf:Q2
Ollama
ollama run hf.co/antirez/deepseek-v4.1-flash-gguf:Q2
Download
hf download antirez/deepseek-v4.1-flash-gguf --include "DeepSeek-V4.1-Flash-Q2.gguf" --local-dir deepseek-v4.1-flash-gguf
MXFP4467.2 GBfile size, 9 parts468.3 GB467.2 + 0.63 + 0.50
Commands
llama.cpp
llama-server -hf antirez/deepseek-v4.1-flash-gguf:MXFP4
Download
hf download antirez/deepseek-v4.1-flash-gguf --include "DeepSeek-V4.1-Flash-1017-MXFP4-experts-*-of-00009.gguf" --local-dir deepseek-v4.1-flash-gguf

MLX for Apple silicon

FormatSizeMemory neededRepo
MLX 2bitTensorFold/DeepSeek-V4.1-Flash-MLX-2bit-MTP222.4 GBfile size223.5 GB222.4 + 0.63 + 0.50Download page
MLXinferencerlabs/DeepSeek-V4.1-Flash-MLX-Q4i434.9 GBfile size436.1 GB434.9 + 0.63 + 0.50Download page
Other conversions (19; 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

23 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
DeepInfra$0.14$0.421Mfp81070.6 s100.0%yesyes
Novita$0.12$0.481Mfp81151.0 s100.0%yesyes
River$0.12$0.481Mnot given1250.7 s99.9%yesyes
Relace$0.12$0.501Mfp41150.7 s100.0%yesyes
Wafer$0.051$0.571Mmxfp41660.3 s100.0%yesyes
InferenceNet$0.10$0.601016kfp81450.3 s100.0%yesyes
Reka$0.12$0.641Mnot given800.3 s99.9%yesyes
CoreWeave$0.13$0.651Mfp81550.7 s100.0%yesyes
Cohere$0.20$0.661Mfp81810.6 s100.0%yesyes
Ionstream$0.20$0.701Mnot given2020.7 s100.0%yesyes
Mancer 2$0.21$0.751Mfp8920.7 s100.0%yesyes
Fireworks$0.21$0.841Mnot given1250.6 s98.2%yesyes
Cloudflare$0.10$0.901Mnot given951.8 s99.1%yesyes
DigitalOcean$0.23$0.901Mnot given1770.7 s100.0%yesyes
OpenInference$0.06$0.901Mfp41450.8 s100.0%yesyes
BaseTen$0.30$1.201Mnot given1820.5 s100.0%yesyes
Modal$0.30$1.201Mnot given781.2 s100.0%yesyes
Nebius$0.30$1.201Mfp81050.7 s100.0%yesyes
Parasail$0.30$1.201Mfp82090.5 s100.0%yesyes
SiliconFlow$0.30$1.201Mfp81241.7 s100.0%yesno
Together$0.30$1.201Mnot given2430.2 s100.0%yesyes
Venice$0.30$1.20977kfp81390.6 s100.0%yesyes
featherless-ai???not given???not givennot 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": "deepseek-ai/DeepSeek-V4.1-Flash: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

from our snapshot of the Hub

Repos that declare this model as their base, most downloaded first. Listing is not a recommendation; unreviewed ones have not been checked.

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.