- Size
- 36.2B parameterscounted from the safetensors header on the Hub
- Licence
- unknown, Apache License 2.0Our reading, check the licence: You can use, change and redistribute the model, including commercially. Keep the licence and any notices, and say which files you changed; you do not have to publish your changes. It includes a patent licence that ends for anyone who sues over patents in the work.
- On the Hub since
- 20 Aug 2025the date the repo was created
- Context length
- 512k tokensfrom the model's config
- Kind
- Base model (no base model declared on the Hub)SeedOssForCausalLM
- On Hugging Face
- ByteDance-Seed/Seed-OSS-36B-Instruct44k downloads in the last 30 days, 506 likes
Scores
- default setting71.5%±2.730 Aug 2026, log
Epoch's name: seed-oss-36b-instruct. Best closed model on this set: gpt-6.2-astra_max, 96.8% (OpenAI). All 341 results
- default setting67.5%±6.328 Aug 2026, log
Epoch's name: seed-oss-36b-instruct. Best closed model on this set: claude-fable-5_high, 100.0% (Anthropic). All 318 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: 2.0 GB, worked out from the model's config (ByteDance-Seed/Seed-OSS-36B-Instruct), 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/Seed-OSS-36B-Instruct-GGUF (converted by unsloth), 7.8k downloads. Sizes are the real files on the Hub, all parts added up.
| Format | Size | Memory needed | Get it |
|---|---|---|---|
| UD-IQ3_XXS | 13.3 GBfile size | 15.8 GB13.3 + 2.0 + 0.50 | Commandsllama.cpp llama-server -hf unsloth/Seed-OSS-36B-Instruct-GGUF:UD-IQ3_XXSOllama ollama run hf.co/unsloth/Seed-OSS-36B-Instruct-GGUF:UD-IQ3_XXSDownload hf download unsloth/Seed-OSS-36B-Instruct-GGUF --include "Seed-OSS-36B-Instruct-UD-IQ3_XXS.gguf" --local-dir Seed-OSS-36B-Instruct-GGUF |
| Q3_K_S | 14.8 GBfile size | 17.3 GB14.8 + 2.0 + 0.50 | Commandsllama.cpp llama-server -hf unsloth/Seed-OSS-36B-Instruct-GGUF:Q3_K_SOllama ollama run hf.co/unsloth/Seed-OSS-36B-Instruct-GGUF:Q3_K_SDownload hf download unsloth/Seed-OSS-36B-Instruct-GGUF --include "Seed-OSS-36B-Instruct-Q3_K_S.gguf" --local-dir Seed-OSS-36B-Instruct-GGUF |
| Q3_K_M | 16.4 GBfile size | 18.9 GB16.4 + 2.0 + 0.50 | Commandsllama.cpp llama-server -hf unsloth/Seed-OSS-36B-Instruct-GGUF:Q3_K_MOllama ollama run hf.co/unsloth/Seed-OSS-36B-Instruct-GGUF:Q3_K_MDownload hf download unsloth/Seed-OSS-36B-Instruct-GGUF --include "Seed-OSS-36B-Instruct-Q3_K_M.gguf" --local-dir Seed-OSS-36B-Instruct-GGUF |
| UD-Q3_K_XL | 16.8 GBfile size | 19.3 GB16.8 + 2.0 + 0.50 | Commandsllama.cpp llama-server -hf unsloth/Seed-OSS-36B-Instruct-GGUF:UD-Q3_K_XLOllama ollama run hf.co/unsloth/Seed-OSS-36B-Instruct-GGUF:UD-Q3_K_XLDownload hf download unsloth/Seed-OSS-36B-Instruct-GGUF --include "Seed-OSS-36B-Instruct-UD-Q3_K_XL.gguf" --local-dir Seed-OSS-36B-Instruct-GGUF |
| IQ4_XS | 18.2 GBfile size | 20.7 GB18.2 + 2.0 + 0.50 | Commandsllama.cpp llama-server -hf unsloth/Seed-OSS-36B-Instruct-GGUF:IQ4_XSOllama ollama run hf.co/unsloth/Seed-OSS-36B-Instruct-GGUF:IQ4_XSDownload hf download unsloth/Seed-OSS-36B-Instruct-GGUF --include "Seed-OSS-36B-Instruct-IQ4_XS.gguf" --local-dir Seed-OSS-36B-Instruct-GGUF |
| IQ4_NL | 19.2 GBfile size | 21.7 GB19.2 + 2.0 + 0.50 | Commandsllama.cpp llama-server -hf unsloth/Seed-OSS-36B-Instruct-GGUF:IQ4_NLOllama ollama run hf.co/unsloth/Seed-OSS-36B-Instruct-GGUF:IQ4_NLDownload hf download unsloth/Seed-OSS-36B-Instruct-GGUF --include "Seed-OSS-36B-Instruct-IQ4_NL.gguf" --local-dir Seed-OSS-36B-Instruct-GGUF |
| Q4_0 | 19.2 GBfile size | 21.7 GB19.2 + 2.0 + 0.50 | Commandsllama.cpp llama-server -hf unsloth/Seed-OSS-36B-Instruct-GGUF:Q4_0Ollama ollama run hf.co/unsloth/Seed-OSS-36B-Instruct-GGUF:Q4_0Download hf download unsloth/Seed-OSS-36B-Instruct-GGUF --include "Seed-OSS-36B-Instruct-Q4_0.gguf" --local-dir Seed-OSS-36B-Instruct-GGUF |
| Q4_K_S | 19.3 GBfile size | 21.8 GB19.3 + 2.0 + 0.50 | Commandsllama.cpp llama-server -hf unsloth/Seed-OSS-36B-Instruct-GGUF:Q4_K_SOllama ollama run hf.co/unsloth/Seed-OSS-36B-Instruct-GGUF:Q4_K_SDownload hf download unsloth/Seed-OSS-36B-Instruct-GGUF --include "Seed-OSS-36B-Instruct-Q4_K_S.gguf" --local-dir Seed-OSS-36B-Instruct-GGUF |
| Q4_K_M | 20.3 GBfile size | 22.8 GB20.3 + 2.0 + 0.50 | Commandsllama.cpp llama-server -hf unsloth/Seed-OSS-36B-Instruct-GGUF:Q4_K_MOllama ollama run hf.co/unsloth/Seed-OSS-36B-Instruct-GGUF:Q4_K_MDownload hf download unsloth/Seed-OSS-36B-Instruct-GGUF --include "Seed-OSS-36B-Instruct-Q4_K_M.gguf" --local-dir Seed-OSS-36B-Instruct-GGUF |
| UD-Q4_K_XL | 20.5 GBfile size | 23.0 GB20.5 + 2.0 + 0.50 | Commandsllama.cpp llama-server -hf unsloth/Seed-OSS-36B-Instruct-GGUF:UD-Q4_K_XLOllama ollama run hf.co/unsloth/Seed-OSS-36B-Instruct-GGUF:UD-Q4_K_XLDownload hf download unsloth/Seed-OSS-36B-Instruct-GGUF --include "Seed-OSS-36B-Instruct-UD-Q4_K_XL.gguf" --local-dir Seed-OSS-36B-Instruct-GGUF |
| Q4_1 | 21.2 GBfile size | 23.7 GB21.2 + 2.0 + 0.50 | Commandsllama.cpp llama-server -hf unsloth/Seed-OSS-36B-Instruct-GGUF:Q4_1Ollama ollama run hf.co/unsloth/Seed-OSS-36B-Instruct-GGUF:Q4_1Download hf download unsloth/Seed-OSS-36B-Instruct-GGUF --include "Seed-OSS-36B-Instruct-Q4_1.gguf" --local-dir Seed-OSS-36B-Instruct-GGUF |
| Q5_K_S | 23.3 GBfile size | 25.8 GB23.3 + 2.0 + 0.50 | Commandsllama.cpp llama-server -hf unsloth/Seed-OSS-36B-Instruct-GGUF:Q5_K_SOllama ollama run hf.co/unsloth/Seed-OSS-36B-Instruct-GGUF:Q5_K_SDownload hf download unsloth/Seed-OSS-36B-Instruct-GGUF --include "Seed-OSS-36B-Instruct-Q5_K_S.gguf" --local-dir Seed-OSS-36B-Instruct-GGUF |
| Q5_K_M | 23.8 GBfile size | 26.3 GB23.8 + 2.0 + 0.50 | Commandsllama.cpp llama-server -hf unsloth/Seed-OSS-36B-Instruct-GGUF:Q5_K_MOllama ollama run hf.co/unsloth/Seed-OSS-36B-Instruct-GGUF:Q5_K_MDownload hf download unsloth/Seed-OSS-36B-Instruct-GGUF --include "Seed-OSS-36B-Instruct-Q5_K_M.gguf" --local-dir Seed-OSS-36B-Instruct-GGUF |
| UD-Q5_K_XL | 23.9 GBfile size | 26.4 GB23.9 + 2.0 + 0.50 | Commandsllama.cpp llama-server -hf unsloth/Seed-OSS-36B-Instruct-GGUF:UD-Q5_K_XLOllama ollama run hf.co/unsloth/Seed-OSS-36B-Instruct-GGUF:UD-Q5_K_XLDownload hf download unsloth/Seed-OSS-36B-Instruct-GGUF --include "Seed-OSS-36B-Instruct-UD-Q5_K_XL.gguf" --local-dir Seed-OSS-36B-Instruct-GGUF |
| Q6_K | 27.6 GBfile size | 30.1 GB27.6 + 2.0 + 0.50 | Commandsllama.cpp llama-server -hf unsloth/Seed-OSS-36B-Instruct-GGUF:Q6_KOllama ollama run hf.co/unsloth/Seed-OSS-36B-Instruct-GGUF:Q6_KDownload hf download unsloth/Seed-OSS-36B-Instruct-GGUF --include "Seed-OSS-36B-Instruct-Q6_K.gguf" --local-dir Seed-OSS-36B-Instruct-GGUF |
| UD-Q6_K_XL | 29.7 GBfile size | 32.2 GB29.7 + 2.0 + 0.50 | Commandsllama.cpp llama-server -hf unsloth/Seed-OSS-36B-Instruct-GGUF:UD-Q6_K_XLOllama ollama run hf.co/unsloth/Seed-OSS-36B-Instruct-GGUF:UD-Q6_K_XLDownload hf download unsloth/Seed-OSS-36B-Instruct-GGUF --include "Seed-OSS-36B-Instruct-UD-Q6_K_XL.gguf" --local-dir Seed-OSS-36B-Instruct-GGUF |
| Q8_0 | 35.8 GBfile size | 38.3 GB35.8 + 2.0 + 0.50 | Commandsllama.cpp llama-server -hf unsloth/Seed-OSS-36B-Instruct-GGUF:Q8_0Ollama ollama run hf.co/unsloth/Seed-OSS-36B-Instruct-GGUF:Q8_0Download hf download unsloth/Seed-OSS-36B-Instruct-GGUF --include "Seed-OSS-36B-Instruct-Q8_0.gguf" --local-dir Seed-OSS-36B-Instruct-GGUF |
| UD-Q8_K_XL | 40.0 GBfile size | 42.5 GB40.0 + 2.0 + 0.50 | Commandsllama.cpp llama-server -hf unsloth/Seed-OSS-36B-Instruct-GGUF:UD-Q8_K_XLOllama ollama run hf.co/unsloth/Seed-OSS-36B-Instruct-GGUF:UD-Q8_K_XLDownload hf download unsloth/Seed-OSS-36B-Instruct-GGUF --include "Seed-OSS-36B-Instruct-UD-Q8_K_XL.gguf" --local-dir Seed-OSS-36B-Instruct-GGUF |
| BF16 | 67.3 GBfile size, 2 parts | 69.8 GB67.3 + 2.0 + 0.50 | Commandsllama.cpp llama-server -hf unsloth/Seed-OSS-36B-Instruct-GGUF:BF16Download hf download unsloth/Seed-OSS-36B-Instruct-GGUF --include "BF16/Seed-OSS-36B-Instruct-BF16-*-of-00002.gguf" --local-dir Seed-OSS-36B-Instruct-GGUF |
MLX for Apple silicon
| Format | Size | Memory needed | Repo |
|---|---|---|---|
| MLX 4bitlmstudio-community/Seed-OSS-36B-Instruct-MLX-4bit | 18.9 GBfile size | 21.4 GB18.9 + 2.0 + 0.50 | Download page |
| MLX 8bitlmstudio-community/Seed-OSS-36B-Instruct-MLX-8bit | 35.8 GBfile size | 38.3 GB35.8 + 2.0 + 0.50 | Download page |
Original weights size from the safetensors header, for vLLM, SGLang, Transformers
| Format | Size | Memory needed | Repo |
|---|---|---|---|
| BF16original weights | 67.3 GBfile size | 69.8 GB67.3 + 2.0 + 0.50 |
hf download ByteDance-Seed/Seed-OSS-36B-Instruct --local-dir Seed-OSS-36B-InstructOther conversions (11; sizes not read)
- QuantTrio/Seed-OSS-36B-Instruct-AWQ AWQ, 176 downloads
- cyankiwi/Seed-OSS-36B-Instruct-AWQ-4bit AWQ, 4bit, 90 downloads
- INC4AI/Seed-OSS-36B-Instruct-int4-AutoRound Other, int4, 62 downloads
- mratsim/Seed-OSS-36B-Instruct-NVFP4 NVFP4, 31 downloads
- cyankiwi/Seed-OSS-36B-Instruct-AWQ-8bit AWQ, 8bit, 28 downloads
- QuantTrio/Seed-OSS-36B-Instruct-GPTQ-Int4 GPTQ, int4, 18 downloads
- mratsim/Seed-OSS-36B-Instruct-FP8-KV8 FP8, 18 downloads
- QuantTrio/Seed-OSS-36B-Instruct-GPTQ-Int3 GPTQ, 17 downloads
- QuantTrio/Seed-OSS-36B-Instruct-GPTQ-Int8 GPTQ, int8, 15 downloads
- JunHowie/Seed-OSS-36B-Instruct-GPTQ-Int8 GPTQ, int8, 11 downloads
- Ataromoku/Seed-OSS-36B-Instruct-exl3 EXL3, 11 downloads
The Hub tags 39 quantised repos as conversions of this model; we read the most downloaded and show the maker's own first.
Run it by API
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, 13:43 UTC. The repo. Found a mistake on this page? Tell us.