BLK · MODEL RELEASE TRACKER

What just dropped.

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

AddedModelParams8GB16GB24GB48GB96GB+
20h agoQwen3.8 27B
qwen
27B✗△~✓✓
20h agoCommand R7B (12-2024)
command-r
8B~✓✓✓✓
20h agoOpenELM 3B Instruct
other
3B✓✓✓✓✓
20h agoSmolVLM Instruct
other
2.25B✓✓✓✓✓
20h agoDeepSeek V2 Lite Chat
deepseek
15.7B✗✓✓✓✓
20h agoOLMo 2 1B Instruct
olmo
1B✓✓✓✓✓
20h agoFalcon 3 3B Instruct
falcon
3B✓✓✓✓✓
20h agoGranite 3.1 2B Instruct
granite
2B✓✓✓✓✓
20h agoQwen2-VL 2B Instruct
qwen
2B✓✓✓✓✓
20h agoTinyLlama 1.1B Chat v1.0
llama
1.1B✓✓✓✓✓
20h agoSmolLM2 360M Instruct
other
360M✓✓✓✓✓
20h agoSmolLM2 135M Instruct
other
135M✓✓✓✓✓
20h agoGemma 2 2B Instruct
gemma
2B✓✓✓✓✓
20h agoGemma 3 270M
gemma
270M✓✓✓✓✓
20h agoQwen 3 1.7B
qwen
1.7B✓✓✓✓✓
20h agoQwen 3 0.6B
qwen
600M✓✓✓✓✓
20h agoparaphrase-multilingual-MiniLM-L12-v2
other
118M✓✓✓✓✓
20h agoall-mpnet-base-v2
other
109M✓✓✓✓✓
20h agoall-MiniLM-L6-v2
other
22M✓✓✓✓✓
20h agoGOT-OCR 2.0
stepfun
580M✓✓✓✓✓
20h agoFlorence-2 Large
other
770M✓✓✓✓✓
20h agoColPali v1.3
gemma
3B✓✓✓✓✓
20h agoSigLIP SO400M (patch14-384)
other
428M✓✓✓✓✓
20h agoStable Diffusion 3.5 Medium
other
2.5B✓✓✓✓✓
20h agoSDXL Turbo
other
2.6B✓✓✓✓✓
20h agoFLUX.1 [schnell]
other
12B△✓✓✓✓
20h agoFLUX.1 [dev]
other
12B△✓✓✓✓
20h agomxbai-rerank-large-v2
other
1.54B✓✓✓✓✓
20h agoJina Reranker v2 Base Multilingual
other
278M✓✓✓✓✓
20h agoMultilingual E5 Large Instruct
other
560M✓✓✓✓✓
20h agoE5 Mistral 7B Instruct
other
7.11B~✓✓✓✓
20h agoGTE ModernBERT Base
other
149M✓✓✓✓✓
20h agoSnowflake Arctic Embed L v2.0
other
568M✓✓✓✓✓
20h agoJina Embeddings v3
other
572M✓✓✓✓✓
20h agomxbai-embed-large-v1
other
335M✓✓✓✓✓
20h agoBGE Large EN v1.5
other
335M✓✓✓✓✓
20h agoNomic Embed Text v1.5
other
137M✓✓✓✓✓
20h agoPiper
other
25M✓✓✓✓✓
20h agoOrpheus 3B 0.1 FT
other
3B✓✓✓✓✓
20h agoF5-TTS
other
336M✓✓✓✓✓
20h agoXTTS v2
other
460M✓✓✓✓✓
20h agoKokoro 82M
other
82M✓✓✓✓✓
20h agoParakeet TDT 0.6B v2
other
600M✓✓✓✓✓
20h agoDistil-Whisper Large v3
other
756M✓✓✓✓✓
20h agoWhisper Small
other
244M✓✓✓✓✓
20h agoWhisper Base
other
74M✓✓✓✓✓
20h agoWhisper Tiny
other
39M✓✓✓✓✓
20h agoMalhajar Mistral 7B Turkish
mistral
7.2B~✓✓✓✓
20h agoRefinedNeuro RN TR R2
llama
8B~✓✓✓✓
20h agoRefinedNeuro RN TR R1
llama
8B~✓✓✓✓
20h agoYTU Turkish Gemma 9B v0.1
gemma
9.2B~✓✓✓✓
20h agoTurkcell LLM 7B v1
mistral
7.4B~✓✓✓✓
20h agoMistral Turkish v2 (brooqs)
mistral
7.2B~✓✓✓✓
20h agoTrendyol LLM Asure 12B
gemma
11.8B△✓✓✓✓
20h agoCosmos Llama 3 8B Turkish
llama
8B~✓✓✓✓
20h agoTrendyol LLM 7B Base v0.1
llama
7B✓✓✓✓✓
20h agoTurkish GPT-2 Large
other
700M✓✓✓✓✓
20h agoVBART Large (Turkish Summarization)
other
400M✓✓✓✓✓
20h agoMihenk LLM v2 35B (Turkish Financial)
other
35B✗✗~✓✓
20h agoOmni 31B Turkish Reasoning
other
31B✗△~✓✓
✓
Comfortable
Q4 fits with KV headroom
~
Tight
Q4 fits, small context
△
Marginal
IQ3 only, expect degradation
✗
Doesn't fit
Won't run usefully
HOW FIT VERDICTS ARE DERIVED

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.