ByteDance Seed Open Models: Model List and Licenses
About ByteDance Seed
ByteDance Seed is the AI research team of ByteDance, the company behind TikTok. It releases language models and research models on its ByteDance-Seed organization on Hugging Face.
What It Releases
- Seed-OSS-36B (August 2025): a 36B language model with base and instruct versions. It is also the base of Nous Research’s Hermes 4.3.
- Diffusion language models: research models that do not generate one token at a time, such as Stable-DiffCoder-8B for code (January 2026) and Cola-DLM (May 2026).
- Theorem proving: BFS-Prover-V2 (7B and 32B), which writes mathematical proofs in the Lean4 proof assistant.
- Research add-ons: such as AHN (Artificial Hippocampus Networks), which adds to Qwen2.5 a way to handle long contexts with fixed-size compressed memory.
Licenses
Seed-OSS and BFS-Prover-V2 are Apache-2.0, and Stable-DiffCoder is MIT. Some, like the AHN add-ons for the 3B Qwen2.5, inherit the original model’s license (Qwen Research License). Read each model’s license text before use. This summary is not legal advice.
Running Them Locally
- There are no official GGUF builds (as of 2026-09-26); for Seed-OSS-36B, quantize a third-party conversion.
- Diffusion language models generate text differently from ordinary language models. llama.cpp and Ollama may not support them.
Sources: the model list and model cards of the ByteDance-Seed organization on Hugging Face; the model card of NousResearch/Hermes-4.3-36B (all as of 2026-09-26).
Our Coverage and Data
Local Model Watch has not published an article on ByteDance Seed’s models yet. The model list below is collected by code from Hugging Face every day. Part of our publisher index; model families link to their family pages.
At a Glance
| Item | Value |
|---|---|
| Hugging Face organization | ByteDance-Seed |
| Models on Hugging Face | 59 |
| Articles on its own models | 0 |
| Articles on derived models | 0 |
Recent Models on Hugging Face
The publisher lists 59 models in total on Hugging Face. Below are up to 15 of the most recent generative models (text, image, video, audio and so on), the scope of this site, per organization. Classifiers, feature extractors, research components and repositories without a model card are omitted, and the publisher’s own quantized builds and helper variants are folded into the row of the original model. Parameter counts are left blank for repositories that pack weights into integer types, where Hugging Face’s count is not the real parameter count.
→ Scroll horizontally to see all columns
| Released | Model | Task | Parameters | License | Official variants |
|---|---|---|---|---|---|
| 2026-06-02 | ByteDance-Seed/TaskMem | — | 31.1B | apache-2.0 |
— |
| 2026-05-19 | ByteDance-Seed/SimArt | — | — | apache-2.0 |
— |
| 2026-05-15 | ByteDance-Seed/Cola-DLM | text generation | — | apache-2.0 |
— |
| 2026-01-15 | ByteDance-Seed/Stable-DiffCoder-8B-Instruct | text generation | 8.3B | mit |
— |
| 2026-01-15 | ByteDance-Seed/Stable-DiffCoder-8B-Base | text generation | 8.3B | mit |
— |
| 2025-10-08 | ByteDance-Seed/AHN-Mamba2-for-Qwen-2.5-Instruct-14B | text generation | 51M | apache-2.0 |
— |
| 2025-10-08 | ByteDance-Seed/AHN-Mamba2-for-Qwen-2.5-Instruct-7B | text generation | 19M | apache-2.0 |
— |
| 2025-10-08 | ByteDance-Seed/AHN-Mamba2-for-Qwen-2.5-Instruct-3B | text generation | 12M | qwen-research |
— |
| 2025-10-08 | ByteDance-Seed/AHN-GDN-for-Qwen-2.5-Instruct-14B | text generation | 61M | apache-2.0 |
— |
| 2025-10-08 | ByteDance-Seed/AHN-GDN-for-Qwen-2.5-Instruct-7B | text generation | 21M | apache-2.0 |
— |
| 2025-10-08 | ByteDance-Seed/AHN-GDN-for-Qwen-2.5-Instruct-3B | text generation | 13M | qwen-research |
— |
| 2025-10-08 | ByteDance-Seed/AHN-DN-for-Qwen-2.5-Instruct-14B | text generation | 51M | apache-2.0 |
— |
| 2025-10-08 | ByteDance-Seed/AHN-DN-for-Qwen-2.5-Instruct-7B | text generation | 18M | apache-2.0 |
— |
| 2025-10-08 | ByteDance-Seed/AHN-DN-for-Qwen-2.5-Instruct-3B | text generation | 12M | qwen-research |
— |
| 2025-10-06 | ByteDance-Seed/BFS-Prover-V2-7B | text generation | 7.6B | apache-2.0 |
— |
From the Hugging Face API. Last changed 2026-09-27 (JST). Collected by code.
Last updated 2026-09-27 (JST). The explanation at the top of this page was written with the help of AI from the primary sources it cites. The tables and lists under “Our Coverage and Data" are assembled by code from our article log and the Hugging Face API.