Up to date
deepseek-ai MIT out 21 months ago
684B parameters (8 of 256 experts per token), text in, text out, 160K tokens of context.
Verified 2026-10-25 Rank #553 in LLMs · score 74.9 method v1 Sources 3 Downloads 30d 4.8k · #546 of 619 tracked
Picked by rule from 32 tracked pieces: the lab's own post, the most shared, a field report, the latest.
Zyphra released ZAYA1-8B, a model that matches DeepSeek-R1 on math benchmarks while using less than 1 billion active parameters.
The 2 picks above, plus these 23. Top is the radar's score: discussion, source authority and the lab's own posts, fading with age.
Hugging Face does not say who downloads a model or with what. These are the proxies: API providers listing it, quantized builds, fine-tunes/adapters/merges, and Spaces that load it.
No API provider serves it, so its downloads are people running it themselves: 6 quantized builds and 66 Spaces.
Top Spaces running it: llamameta/DeepSeek-R1-Chat-Assistant-Web-Search ♥ 260 · KBaba7/Quant ♥ 51 · ruslanmv/DeepSeek-R1-Chatbot ♥ 36 · Sarath0x8f/Document-QA-bot ♥ 33
Weights only, an estimate: leave room for the KV cache. Method
About 411 GB at INT4, 821 GB at FP8, 1.64 TB at BF16: weights only, so leave room for the KV cache. Every expert must be resident, so total parameters count.
Estimate params × bytes per param × 1.2 · how we estimate
Weights, every dated article, and repo commits on one rail.
Teal is the weights (2025-01-20), amber triangles are 10 dated articles (each links to it), grey ticks are repo commits.
18 derivatives on the Hub. Which build to run, by runtime: the most downloaded in each class.
From the Hugging Face model tree; quantized builds in teal, fine-tunes, merges and adapters in greys.
Everything else, folded: the full tables, every source.
Labels are evidence, not verdicts: measured a third-party leaderboard row for this model; author-reported a number the model author published in the card or eval files; structural a deterministic signal from the name, config or chat template, or a decision-model estimate. Each chip links to its source. All use cases.
| Pipeline tag | text-generation |
|---|---|
| Library | transformers |
| Architectures | DeepseekV3ForCausalLM |
| Model type | deepseek_v3 |
| Dtype | bfloat16 |
| Params by dtype | BF16: 3,918,786,560 · F32: 15,104 · F8_E4M3: 680,571,043,840 |
| Size band | XXL |
| MoE | yes |
| Experts | 256 |
| Experts per token | 8 |
| Context length | 163,840 (config.json max_position_embeddings) |
| Papers | arXiv:2501.12948 |
| License | MIT (mit) |
| Hub snapshot | 2026-10-25 |
vLLM: supported · DeepseekV3ForCausalLM · 7050ef881ff0835f6890a47db1859a24c65481d6SGLang: supported · DeepseekV3ForCausalLM · 773cb938fe9484a8f7e7e46fb1b1fe96412c59ed
vLLM registry (7050ef881ff0835f6890a47db1859a24c65481d6) · SGLang registry (773cb938fe9484a8f7e7e46fb1b1fe96412c59ed) · Matched by architecture class name against each project's registry on main; 'supported' means the architecture is registered, not that this checkpoint was tested.
Weights-only memory for DeepSeek-R1-Zero: params × bytes/param × 1.2 overhead. KV cache and activations are extra. How this is computed.
| BF16 (2 bytes/param) | 1.64 TB · Cluster (≥640 GB) |
|---|---|
| FP8 (1 byte/param) | 821 GB · Cluster (≥640 GB) |
| INT4 (0.5 bytes/param) | 411 GB · Multi-GPU node (<640 GB) |
Weights only (params × bytes/param × 1.2 overhead); KV cache and activations extra. MoE: all experts must be resident, so total parameters count.
quantized: 6finetune: 11adapter: 1
| Variant | Relation | Quant | Params | Band | License | Author | Likes | Downloads |
|---|---|---|---|---|---|---|---|---|
| unsloth/DeepSeek-R1-Zero-GGUF | quantized | GGUF | — | — | mit | unsloth | 27 | 40.8k |
| DevQuasar-3/deepseek-ai.DeepSeek-R1-Zero-GGUF | quantized | GGUF | — | — | — | DevQuasar-3 | 2 | 3.9k |
| unsloth/DeepSeek-R1-Zero | quantized | — | 685B | XXL | mit | unsloth | 6 | 2.1k |
| Daemontatox/Zireal-0 | finetune | — | 684B | XXL | apache-2.0 | Daemontatox | 1 | 2.2k |
| opensourcerelease/DeepSeek-R1-Zero-bf16 | finetune | -bf16 | 684B | XXL | mit | opensourcerelease | 0 | 587 |
| gghfez/DeepSeek-R1-Zero-256x21B-BF16 | finetune | -BF16 | — | — | mit | gghfez | 0 | 422 |
| gghfez/DeepSeek-R1-Zero-IQ2_KS | quantized | GGUF | — | — | mit | gghfez | 0 | 292 |
| gghfez/DeepSeek-R1-Zero-IQ3_KS | quantized | GGUF | — | — | mit | gghfez | 0 | 288 |
| adamo1139/DeepSeek-R1-Zero-AWQ | quantized | AWQ | 671B | XXL | mit | adamo1139 | 0 | 213 |
| mradermacher/DeepSeek-R1-Zero-GGUF | finetune | GGUF | — | — | mit | mradermacher | 1 | 0 |
| mradermacher/DeepSeek-R1-Zero-i1-GGUF | finetune | GGUF | — | — | mit | mradermacher | 1 | 0 |
| osllmai-community/DeepSeek-R1-Zero-GGUF | finetune | GGUF | — | — | mit | osllmai-community | 0 | 0 |
| Oi2qtop/sigma | finetune | — | — | — | — | Oi2qtop | 0 | 0 |
| Elchapofx28/Girl | finetune | — | — | — | afl-3.0 | Elchapofx28 | 0 | 0 |
| Xacales/code1 | finetune | — | — | — | gfdl | Xacales | 0 | 0 |
| saha78rus/Realistic | finetune | — | — | — | apache-2.0 | saha78rus | 0 | 0 |
| Etan498/inferno | finetune | — | — | — | unknown | Etan498 | 0 | 0 |
| shynword/CHATGPT313 | adapter | — | — | — | apache-2.0 | shynword | 0 | 0 |
| Date | Commit | SHA |
|---|---|---|
| 2025-03-27 | Small fix | 7223428 |
| 2025-02-24 | Update bib info | 9d8dd91 |
| 2025-02-09 | Update tokenizer_config.json | 978b97c |
| 2025-02-01 | Update README.md | b3c33b0 |
| 2025-01-26 | Update README.md | d9c339c |
| 2025-01-23 | Update README.md | b039779 |
| 2025-01-22 | Update tokenizer_config.json | 19de63b |
| 2025-01-22 | Update tokenizer_config.json | d3cf750 |
| 2025-01-21 | Update README.md (#12) | 9cc4dfb |
| 2025-01-21 | add library name & auto-tag (#10) | 57b1029 |
| 2025-01-21 | Add generation_config.json | 7c4fde6 |
| 2025-01-20 | Release DeepSeek-R1 | cfa4541 |
| 2025-01-20 | Add files using upload-large-folder tool | 3a13570 |
| 2025-01-20 | Add files using upload-large-folder tool | 4a087ad |
| 2025-01-20 | Add files using upload-large-folder tool | 5f6c31f |
| 2025-01-20 | Add files using upload-large-folder tool | 6c57e3e |
| 2025-01-20 | Add files using upload-large-folder tool | acd6df8 |
| 2025-01-20 | Add files using upload-large-folder tool | d511d25 |
| 2025-01-20 | Add files using upload-large-folder tool | 16842fd |
| 2025-01-20 | Add files using upload-large-folder tool | 8c9cab6 |
| 2025-01-20 | Add files using upload-large-folder tool | 6b06717 |
| 2025-01-20 | Add files using upload-large-folder tool | 2aafb87 |
| 2025-01-20 | Add files using upload-large-folder tool | 60c0395 |
| 2025-01-20 | Add files using upload-large-folder tool | fb5c8d8 |
| 2025-01-20 | Add files using upload-large-folder tool | e30d997 |
| 2025-01-20 | Add files using upload-large-folder tool | a0ba933 |
| 2025-01-20 | initial commit | 24dee08 |
Ranked by source signals (HN, Lobsters, authority domains, official posts) and recency. Summaries are generated from a stored copy of the article text; every sentence cites a span of that text, and nothing is shown that failed the citation check.
arXiv links found in the model card and its tags. "Peer-reviewed" is shown only with evidence: a published venue in OpenAlex (measured) or the authors' own journal reference on arXiv (reported).
Hub snapshot: 2026-10-25 · HF API: 2026-10-24 21:40 UTC · Providers: 2026-10-24 21:44 UTC · Articles: 2026-10-24 21:46 UTC · Summary: 2026-10-07 15:12 UTC · Last verified 2026-10-25
Email alerts are not open yet; the feeds above are the same data. See feeds for the item schema.
