SAVRN Model Hub
Open-Weight Models
An open-weight model is an AI model whose trained weights are published for anyone to download. The weights are what the model learned in training. With a copy of them you can run the model on hardware you control and train it further on your own data.
Open weights are not the same as open source. Many publishers release the weights without the training data or code, and the license sets what you may do with the model. This library puts each model's full card, architecture, files, license and published evaluations on one page.
Updated 2026-10-23 · How the library is built
12,579 models, sorted by most downloaded.
Epoch checkpoints of the Qwen3.8-27B DFlash2 MoS line (project MoS, branch ryan/mos-improve, experiments/dflash2/qwen3.8-27b-100k/). Started 2026-09-22, when the general archive repo ryan-0608/MoS-Aurora-Experiment-Archive reached the Hugging Face 20,000-file limit. Serving protocol: sglang main f5866545, TP2, DFLASH 16 draft tokens, fixed-500 prompts, max 128 new tokens, greedy, concurrency 1, 500/500 completed. Same protocol as the dense and K4-jitter references. Reference arms (model-only, no optimizer) are in ryan-0608/MoS-Aurora-Experiment-Archive under dflash227btraj20260915/main-800k/: dense/epoch1, dense/epoch2, mos/epoch1 (K4 shared+1% jitter, expert LR 1e-4; serving 5.1501…
Checkpoints of the Qwen3-8B MoS DFlash draft models (clean-800K E5 follow-up arms, 8B self-generated corpus dataclean-8b-20260912) and the warm-start inputs they were trained from. Project: MoS (mixture of layer-local expert MLPs added to a shared MLP in each DFlash draft layer). Milestones: 13,325 steps = 1 epoch (E1); 26,650 E2; 39,975 E3; 53,300 E4; 66,625 E5; 6,662 / 19,986 / 33,310 / 46,634 / 59,958 are half-epoch saves. Results, configs and logs: GitHub MoS repository, experiments/aurora/qwen3-8b-800k-e5/bal01/README.md (§10). Archive records (file counts and byte totals, verified against the local copies before deletion): experiments/aurora/qwen3-8b-800k-e5/archive/.
MoSE3 is a feed-forward model that predicts world-space SE(3) motion for every pixel of a monocular RGB video, generalizing across rigid, articulated and deformable objects. The code repository has the full inference and visualization scripts. The weights are released under CC BY-NC 4.0 (non-commercial). They include the weights of π³, which are released under the same license.
A small, unprofessional fine-tune of Qwen 2.5, made to have a slightly different personality. Based on huihui-ai/Qwen2.5-0.5B-Instruct-abliterated-v3. Mochi prioritizes having multiple different variants of the base model for different purposes, always using 0.5B parameters. It is only made for fun, and has no actual use if you're looking for something powerful or coherent. - No system prompt required. The personality is now baked in by default. You can still pass a system prompt to adjust behavior, but the Mochi vibe works without one. - Cleaner dataset. 868 examples, all without system prompts, so the model doesn't depend on one to behave correctly. - Factual accuracy is degraded compared…
static quants of https://huggingface.co/wonkaee27/Mochi-4B
This model is a fine-tuned version of ModernBERT optimized for logical reasoning, deductive analysis, and structure-based token prediction. It relies on the ModernBertForMaskedLM architecture, making it highly effective at filling in missing contextual logic tokens (fill-mask). - As an encoder-based Masked Language Model, it is not designed for long-form generative text (like ChatGPT). It excels at predicting masked tokens inside structured prompts.
This is an ONNX version of JohnGorri/ModernBERT-Large-Instruct-Logician-v0.5. It was automatically converted and uploaded using this Hugging Face Space. See the pipeline documentation for fill-mask: https://huggingface.co/docs/transformers.js/api/pipelines#modulepipelines.FillMaskPipeline This model is a fine-tuned version of ModernBERT optimized for logical reasoning, deductive analysis, and structure-based token prediction. It relies on the ModernBertForMaskedLM architecture, making it highly effective at filling in missing contextual logic tokens (fill-mask). - As an encoder-based Masked Language Model, it is not designed for long-form generative text (like ChatGPT). It excels at…
180B total / ~78B active per request — giant knowledge, agile inference. - 64 domain experts, zero overlap — every training example maps to exactly one expert. Auditable by construction. - Fully auditable data pipeline — raw → tagged → tokenized corpora, code included, in Morpho-72B-MoE-Data. - Resume-first engineering — every expert checkpoint (weights + loss + step) is versioned in Morpho-72B-MoE-Checkpoints. Any box can pick up exactly where the last one died. - 128K context via RoPE YaRN for long documents, codebases, and books. Eight sectors — Finance, Law, Medicine, Tech, Science, Business, Arts, Humanities — each with 8 specialists. Full roster with per-expert data counts is logged…
Model Collections
Hand-picked starting points, each with the reason it exists.
Collection · 4 entries
Embedding models for retrieval
Sentence and document embedding models used to build retrieval systems. Dimension and sequence length matter more than size here, and both come from the publisher.
Collection · 4 entries
Models that fit on one accelerator
Models whose publisher-reported parameter count puts them within reach of a single accelerator at common precisions. Memory needed depends on precision and serving configuration, so treat the parameter count as the starting point, not the answer.
Collection · 6 entries
Open-weight text models worth knowing
Widely used open-weight language models, chosen because each one is a distinct family rather than a variant of the one above it. Selection, not a ranking.
Collection · 3 entries
Speech and audio models
Recognition and synthesis models, grouped so the two directions are easy to compare.
Open-Weight Models Explained
What is an open-weight model?
An AI model whose trained weights are published for anyone to download, so it can be run, tested and fine-tuned on hardware the user controls.
Is an open-weight model the same as open source?
Not always. Open weights means the trained model can be downloaded. Open source usually also means the training code and data are available and the license allows broad reuse. Many open-weight models release the weights only.
Can I use an open-weight model commercially?
It depends on the license. Apache 2.0 and MIT allow commercial use. Other licenses limit it, for example to non-commercial use or below a set number of users. Every model page here shows its license.
How much memory does an open-weight model need?
About two bytes per parameter at 16-bit precision, so a 7-billion-parameter model needs roughly 14 GB for its weights, plus memory for the context it processes. Each model page lists its parameter count and the size of its files.
Related SAVRN Research
The hub sits beside SAVRN's market data and infrastructure research: what models cost to run, and what it takes to run them.
SAVRN Index
What open models cost to run
The same open-weight model priced by every host that serves it, per million tokens.
Research Hub
Data center trackers and maps
Moratoriums, permits, power, water and capital behind the facilities that run these models.
Method
How the Model Hub is built
Sources, evidence labels, refresh behaviour, and the limits of every comparison here.