Newest local AI models, date-sorted. Every row carries quick fit verdicts for the four VRAM classes operators ask about — so you know in one glance whether to bother downloading a model before it starts loading. 60 models indexed.
For broader ecosystem news see /pulse. For the recommendation engine see /choose-my-gpu.
60 models shown · newest first
| Added | Model | Params | 8GB | 16GB | 24GB | 48GB | 96GB+ |
|---|---|---|---|---|---|---|---|
| 1w ago | Muse Glimmer 30B other · released 2026-08-10 | 30B | ✗ | △ | ~ | ✓ | ✓ |
| 1mo ago | Nemotron 3 Nano Omni 33B other | 33B | ✗ | △ | ~ | ✓ | ✓ |
| 1mo ago | Ornith 1.0 35B other · released 2026-06-25 | 35B | ✗ | ✗ | ~ | ✓ | ✓ |
| 1mo ago | Qwen3.6 35B-A3B qwen | 35B | ✗ | ✗ | ~ | ✓ | ✓ |
| 1mo ago | Qwen3.5 35B-A3B qwen | 35B | ✗ | ✗ | ~ | ✓ | ✓ |
| 1mo ago | Laguna XS 2.1 other · released 2026-07-02 | 33B | ✗ | △ | ~ | ✓ | ✓ |
| 1mo ago | North Mini Code 1.0 other · released 2026-06-09 | 30B | ✗ | △ | ~ | ✓ | ✓ |
| 1mo ago | GLM-4.7-Flash glm · released 2026-01-19 | 31B | ✗ | △ | ~ | ✓ | ✓ |
| 1mo ago | Qwen3 Coder 30B-A3B qwen | 30B | ✗ | △ | ~ | ✓ | ✓ |
| 1mo ago | Granite 4.1 30B granite | 30B | ✗ | △ | ~ | ✓ | ✓ |
| 1mo ago | Qwen3.6 27B qwen · released 2026-04-22 | 27B | ✗ | △ | ~ | ✓ | ✓ |
| 1mo ago | Qwen3.5 27B qwen | 27B | ✗ | △ | ~ | ✓ | ✓ |
| 1mo ago | Gemma 4 26B-A4B gemma | 26B | ✗ | △ | ✓ | ✓ | ✓ |
| 1mo ago | Dolphin 3.0 8B dolphin | 8B | ~ | ✓ | ✓ | ✓ | ✓ |
| 1mo ago | Mellum2 12B-A2.5B other | 12.15B | △ | ✓ | ✓ | ✓ | ✓ |
| 1mo ago | Ministral 3 14B mistral | 14B | △ | ✓ | ✓ | ✓ | ✓ |
| 1mo ago | GPT-OSS 20B other | 20.9B | ✗ | ~ | ✓ | ✓ | ✓ |
| 1mo ago | Ornith 1.0 9B other · released 2026-06-25 | 9B | ~ | ✓ | ✓ | ✓ | ✓ |
| 1mo ago | Qwen3.5 9B qwen · released 2026-03-02 | 9B | ~ | ✓ | ✓ | ✓ | ✓ |
| 1mo ago | Gemma 4 12B gemma · released 2026-04-02 | 12B | △ | ✓ | ✓ | ✓ | ✓ |
| 1mo ago | LFM2.5-230M other · released 2026-06-25 | 230M | ✓ | ✓ | ✓ | ✓ | ✓ |
| 1mo ago | LongCat-2.0 other · released 2026-07-05 | 1600B | ✗ | ✗ | ✗ | ✗ | ✗ |
| 1mo ago | Hunyuan 3.0 (Hy3) hunyuan · released 2026-07-06 | 295B | ✗ | ✗ | ✗ | ✗ | ✗ |
| 1mo ago | Granite 4.1 30B Instruct granite · released 2026-04-29 | 28.87B | ✗ | △ | ~ | ✓ | ✓ |
| 1mo ago | Granite 4.1 8B Instruct granite · released 2026-04-29 | 8.79B | ~ | ✓ | ✓ | ✓ | ✓ |
| 1mo ago | Granite 4.1 3B Instruct granite · released 2026-04-29 | 3.4B | ✓ | ✓ | ✓ | ✓ | ✓ |
| 1mo ago | Kimi K2.7-Code moonshot · released 2026-06-12 | 1000B | ✗ | ✗ | ✗ | ✗ | ✗ |
| 1mo ago | VibeThinker-3B qwen · released 2026-06-12 | 3B | ✓ | ✓ | ✓ | ✓ | ✓ |
| 1mo ago | Nemotron 3 Ultra (550B-A55B) other · released 2026-06-04 | 550B | ✗ | ✗ | ✗ | ✗ | ✗ |
| 1mo ago | MiniMax-M3 other · released 2026-06-02 | 428B | ✗ | ✗ | ✗ | ✗ | ✗ |
| 1mo ago | GLM-5.2 glm · released 2026-06-16 | 753B | ✗ | ✗ | ✗ | ✗ | ✗ |
| 2mo ago | paraphrase-multilingual-MiniLM-L12-v2 other | 118M | ✓ | ✓ | ✓ | ✓ | ✓ |
| 2mo ago | all-mpnet-base-v2 other | 109M | ✓ | ✓ | ✓ | ✓ | ✓ |
| 2mo ago | all-MiniLM-L6-v2 other | 22M | ✓ | ✓ | ✓ | ✓ | ✓ |
| 2mo ago | Command R7B (12-2024) command-r | 8B | ~ | ✓ | ✓ | ✓ | ✓ |
| 2mo ago | OpenELM 3B Instruct other | 3B | ✓ | ✓ | ✓ | ✓ | ✓ |
| 2mo ago | SmolVLM Instruct other | 2.25B | ✓ | ✓ | ✓ | ✓ | ✓ |
| 2mo ago | DeepSeek V2 Lite Chat deepseek | 15.7B | ✗ | ✓ | ✓ | ✓ | ✓ |
| 2mo ago | OLMo 2 1B Instruct olmo | 1B | ✓ | ✓ | ✓ | ✓ | ✓ |
| 2mo ago | Falcon 3 3B Instruct falcon | 3B | ✓ | ✓ | ✓ | ✓ | ✓ |
| 2mo ago | Granite 3.1 2B Instruct granite | 2B | ✓ | ✓ | ✓ | ✓ | ✓ |
| 2mo ago | Qwen2-VL 2B Instruct qwen | 2B | ✓ | ✓ | ✓ | ✓ | ✓ |
| 2mo ago | TinyLlama 1.1B Chat v1.0 llama | 1.1B | ✓ | ✓ | ✓ | ✓ | ✓ |
| 2mo ago | SmolLM2 360M Instruct other | 360M | ✓ | ✓ | ✓ | ✓ | ✓ |
| 2mo ago | SmolLM2 135M Instruct other | 135M | ✓ | ✓ | ✓ | ✓ | ✓ |
| 2mo ago | Gemma 2 2B Instruct gemma | 2B | ✓ | ✓ | ✓ | ✓ | ✓ |
| 2mo ago | Gemma 3 270M gemma | 270M | ✓ | ✓ | ✓ | ✓ | ✓ |
| 2mo ago | Qwen 3 1.7B qwen | 1.7B | ✓ | ✓ | ✓ | ✓ | ✓ |
| 2mo ago | Qwen 3 0.6B qwen | 600M | ✓ | ✓ | ✓ | ✓ | ✓ |
| 2mo ago | GOT-OCR 2.0 stepfun | 580M | ✓ | ✓ | ✓ | ✓ | ✓ |
| 2mo ago | Florence-2 Large other | 770M | ✓ | ✓ | ✓ | ✓ | ✓ |
| 2mo ago | ColPali v1.3 gemma | 3B | ✓ | ✓ | ✓ | ✓ | ✓ |
| 2mo ago | SigLIP SO400M (patch14-384) other | 428M | ✓ | ✓ | ✓ | ✓ | ✓ |
| 2mo ago | Stable Diffusion 3.5 Medium other | 2.5B | ✓ | ✓ | ✓ | ✓ | ✓ |
| 2mo ago | SDXL Turbo other | 2.6B | ✓ | ✓ | ✓ | ✓ | ✓ |
| 2mo ago | FLUX.1 [schnell] other | 12B | △ | ✓ | ✓ | ✓ | ✓ |
| 2mo ago | FLUX.1 [dev] other | 12B | △ | ✓ | ✓ | ✓ | ✓ |
| 2mo ago | mxbai-rerank-large-v2 other | 1.54B | ✓ | ✓ | ✓ | ✓ | ✓ |
| 2mo ago | Jina Reranker v2 Base Multilingual other | 278M | ✓ | ✓ | ✓ | ✓ | ✓ |
| 2mo ago | Multilingual E5 Large Instruct other | 560M | ✓ | ✓ | ✓ | ✓ | ✓ |
Quick footprint estimate at Q4_K_M: params × 0.6 GB + 1.5 GB runtime overhead. Comfortable means the rig has ≥1.4× headroom for KV cache and multi-turn context. Tight means it fits but you'll bump the ceiling on long context. Marginal means only aggressive (IQ3 / IQ2) quants work, with quality degradation. Doesn't fit means weights alone won't load without RAM offload, which crushes tok/s. Frontier-class models (400B+) render as no-fit on every single-rig VRAM class — they need multi-GPU or cloud.