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+ |
|---|---|---|---|---|---|---|---|
| 2d ago | Kimi K2.7-Code moonshot · released 2026-06-12 | 1000B | ✗ | ✗ | ✗ | ✗ | ✗ |
| 2d ago | VibeThinker-3B qwen · released 2026-06-12 | 3B | ✓ | ✓ | ✓ | ✓ | ✓ |
| 2d ago | Nemotron 3 Ultra (550B-A55B) other · released 2026-06-04 | 550B | ✗ | ✗ | ✗ | ✗ | ✗ |
| 2d ago | MiniMax-M3 other · released 2026-06-02 | 428B | ✗ | ✗ | ✗ | ✗ | ✗ |
| 2d ago | GLM-5.2 glm · released 2026-06-16 | 753B | ✗ | ✗ | ✗ | ✗ | ✗ |
| 1mo ago | paraphrase-multilingual-MiniLM-L12-v2 other | 118M | ✓ | ✓ | ✓ | ✓ | ✓ |
| 1mo ago | all-mpnet-base-v2 other | 109M | ✓ | ✓ | ✓ | ✓ | ✓ |
| 1mo ago | all-MiniLM-L6-v2 other | 22M | ✓ | ✓ | ✓ | ✓ | ✓ |
| 1mo ago | Command R7B (12-2024) command-r | 8B | ~ | ✓ | ✓ | ✓ | ✓ |
| 1mo ago | OpenELM 3B Instruct other | 3B | ✓ | ✓ | ✓ | ✓ | ✓ |
| 1mo ago | SmolVLM Instruct other | 2.25B | ✓ | ✓ | ✓ | ✓ | ✓ |
| 1mo ago | DeepSeek V2 Lite Chat deepseek | 15.7B | ✗ | ✓ | ✓ | ✓ | ✓ |
| 1mo ago | OLMo 2 1B Instruct olmo | 1B | ✓ | ✓ | ✓ | ✓ | ✓ |
| 1mo ago | Falcon 3 3B Instruct falcon | 3B | ✓ | ✓ | ✓ | ✓ | ✓ |
| 1mo ago | Granite 3.1 2B Instruct granite | 2B | ✓ | ✓ | ✓ | ✓ | ✓ |
| 1mo ago | Qwen2-VL 2B Instruct qwen | 2B | ✓ | ✓ | ✓ | ✓ | ✓ |
| 1mo ago | TinyLlama 1.1B Chat v1.0 llama | 1.1B | ✓ | ✓ | ✓ | ✓ | ✓ |
| 1mo ago | SmolLM2 360M Instruct other | 360M | ✓ | ✓ | ✓ | ✓ | ✓ |
| 1mo ago | SmolLM2 135M Instruct other | 135M | ✓ | ✓ | ✓ | ✓ | ✓ |
| 1mo ago | Gemma 2 2B Instruct gemma | 2B | ✓ | ✓ | ✓ | ✓ | ✓ |
| 1mo ago | Gemma 3 270M gemma | 270M | ✓ | ✓ | ✓ | ✓ | ✓ |
| 1mo ago | Qwen 3 1.7B qwen | 1.7B | ✓ | ✓ | ✓ | ✓ | ✓ |
| 1mo ago | Qwen 3 0.6B qwen | 600M | ✓ | ✓ | ✓ | ✓ | ✓ |
| 1mo ago | GOT-OCR 2.0 stepfun | 580M | ✓ | ✓ | ✓ | ✓ | ✓ |
| 1mo ago | Florence-2 Large other | 770M | ✓ | ✓ | ✓ | ✓ | ✓ |
| 1mo ago | ColPali v1.3 gemma | 3B | ✓ | ✓ | ✓ | ✓ | ✓ |
| 1mo ago | SigLIP SO400M (patch14-384) other | 428M | ✓ | ✓ | ✓ | ✓ | ✓ |
| 1mo ago | Stable Diffusion 3.5 Medium other | 2.5B | ✓ | ✓ | ✓ | ✓ | ✓ |
| 1mo ago | SDXL Turbo other | 2.6B | ✓ | ✓ | ✓ | ✓ | ✓ |
| 1mo ago | FLUX.1 [schnell] other | 12B | △ | ✓ | ✓ | ✓ | ✓ |
| 1mo ago | FLUX.1 [dev] other | 12B | △ | ✓ | ✓ | ✓ | ✓ |
| 1mo ago | mxbai-rerank-large-v2 other | 1.54B | ✓ | ✓ | ✓ | ✓ | ✓ |
| 1mo ago | Jina Reranker v2 Base Multilingual other | 278M | ✓ | ✓ | ✓ | ✓ | ✓ |
| 1mo ago | Multilingual E5 Large Instruct other | 560M | ✓ | ✓ | ✓ | ✓ | ✓ |
| 1mo ago | E5 Mistral 7B Instruct other | 7.11B | ~ | ✓ | ✓ | ✓ | ✓ |
| 1mo ago | GTE ModernBERT Base other | 149M | ✓ | ✓ | ✓ | ✓ | ✓ |
| 1mo ago | Snowflake Arctic Embed L v2.0 other | 568M | ✓ | ✓ | ✓ | ✓ | ✓ |
| 1mo ago | Jina Embeddings v3 other | 572M | ✓ | ✓ | ✓ | ✓ | ✓ |
| 1mo ago | mxbai-embed-large-v1 other | 335M | ✓ | ✓ | ✓ | ✓ | ✓ |
| 1mo ago | BGE Large EN v1.5 other | 335M | ✓ | ✓ | ✓ | ✓ | ✓ |
| 1mo ago | Nomic Embed Text v1.5 other | 137M | ✓ | ✓ | ✓ | ✓ | ✓ |
| 1mo ago | Piper other | 25M | ✓ | ✓ | ✓ | ✓ | ✓ |
| 1mo ago | Orpheus 3B 0.1 FT other | 3B | ✓ | ✓ | ✓ | ✓ | ✓ |
| 1mo ago | F5-TTS other | 336M | ✓ | ✓ | ✓ | ✓ | ✓ |
| 1mo ago | XTTS v2 other | 460M | ✓ | ✓ | ✓ | ✓ | ✓ |
| 1mo ago | Kokoro 82M other | 82M | ✓ | ✓ | ✓ | ✓ | ✓ |
| 1mo ago | Parakeet TDT 0.6B v2 other | 600M | ✓ | ✓ | ✓ | ✓ | ✓ |
| 1mo ago | Distil-Whisper Large v3 other | 756M | ✓ | ✓ | ✓ | ✓ | ✓ |
| 1mo ago | Whisper Small other | 244M | ✓ | ✓ | ✓ | ✓ | ✓ |
| 1mo ago | Whisper Base other | 74M | ✓ | ✓ | ✓ | ✓ | ✓ |
| 1mo ago | Whisper Tiny other | 39M | ✓ | ✓ | ✓ | ✓ | ✓ |
| 1mo ago | Sarvam 30B other | 30B | ✗ | △ | ~ | ✓ | ✓ |
| 1mo ago | Typhoon S ThaiLLM 8B Instruct Research Preview other | 8B | ~ | ✓ | ✓ | ✓ | ✓ |
| 1mo ago | OpenThaiGPT 1.5 7B Instruct other | 7B | ✓ | ✓ | ✓ | ✓ | ✓ |
| 1mo ago | PhoGPT 4B other | 3.7B | ✓ | ✓ | ✓ | ✓ | ✓ |
| 1mo ago | PhoGPT 4B Chat other | 3.7B | ✓ | ✓ | ✓ | ✓ | ✓ |
| 1mo ago | Vikhr Qwen 2.5 0.5B Instruct other | 500M | ✓ | ✓ | ✓ | ✓ | ✓ |
| 1mo ago | Dostoevsky Doesn't Write It GPT2 other | 175M | ✓ | ✓ | ✓ | ✓ | ✓ |
| 1mo ago | mGPT 13B other | 13B | △ | ✓ | ✓ | ✓ | ✓ |
| 1mo ago | mGPT 1.3B Mongol other | 1.3B | ✓ | ✓ | ✓ | ✓ | ✓ |
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.