| DeepSeek V4 Flash |
https://www.reddit.com/r/LocalLLaMA/comments/1vjklwo/deepseek_v4_flash_0731_hits_827_on_terminalbench/ |
provider |
integrated |
docs/tools/providers/deepseek.md |
A flash version of the DeepSeek model series. |
| Supraelegans-500K |
https://www.reddit.com/r/LocalLLaMA/comments/1vk3xpb/new_model_supraelegans500k/ |
tool |
integrated |
docs/tools/ai_knowledge/supraelegans.md |
A new LLM model release. |
| OMLab-VLX-Seek-15-10B |
https://www.reddit.com/r/LocalLLaMA/comments/1vkaypz/omlabvlxseek1510b_hugging_face/ |
tool |
integrated |
docs/tools/ai_knowledge/omlab-vlx-seek.md |
A vision-language model available on Hugging Face. |
| EndlessFrontier-BigBang-V1 |
https://www.reddit.com/r/LocalLLaMA/comments/1vk1p9s/endlessfrontierbigbangv1_qwen_35_finetunes/ |
tool |
integrated |
docs/tools/ai_knowledge/endlessfrontier.md |
Finetunes based on Qwen 3.5. |
| MiniMax-H3 |
https://www.reddit.com/r/LocalLLaMA/comments/1vkdatu/minimax_h3_a_new_openweight_video_model_live_in/ |
tool |
integrated |
docs/tools/providers/minimax.md |
An open-weight video generation model. |
| WorldClaw |
https://www.reddit.com/r/LocalLLaMA/comments/1vjnqmh/tencent_announce_worldclaw/ |
tool |
integrated |
docs/tools/agents/worldclaw.md |
An AI agentic world-model framework announced by Tencent. |
| WeatherNext 2 |
https://www.reddit.com/r/LocalLLaMA/comments/1vjwwrs/open_model_google_weather_next_2/ |
tool |
integrated |
docs/tools/ai_knowledge/weathernext.md |
An open weather forecasting model from Google DeepMind. |
| Lophius |
https://www.reddit.com/r/LocalLLaMA/comments/1vjt4vi/lophius_a_workbench_for_language_model_research/ |
tool |
integrated |
docs/tools/development_ops/lophius.md |
A workbench designed for language model research. |
| KLQ |
https://www.reddit.com/r/LocalLLaMA/comments/1vk2n2k/klq_trainingfree_measured_rotation_quantization/ |
framework |
integrated |
docs/tools/infrastructure/klq.md |
A training-free measured rotation quantization method. |