{"id":3547,"date":"2026-09-25T11:21:30","date_gmt":"2026-09-25T02:21:30","guid":{"rendered":"https:\/\/localmodelwatch.tsuchitsuchi.com\/vram-guide-16gb-en\/"},"modified":"2026-09-26T04:11:19","modified_gmt":"2026-09-25T19:11:19","slug":"vram-guide-16gb-en","status":"publish","type":"page","link":"https:\/\/localmodelwatch.tsuchitsuchi.com\/en\/vram-guide-16gb-en\/","title":{"rendered":"Local Models That Run in 16GB of VRAM (RTX 5060 Ti 16GB \/ 4060 Ti 16GB, etc.)"},"content":{"rendered":"

Every model covered by Local Model Watch that fits in 16GB<\/strong> of GPU memory, with the best build that fits. The “Best build for 16GB” column is the largest (highest-quality) quantization or precision whose estimated memory stays within 16GB \u2014 often better than the smallest build listed in the VRAM quick reference<\/a>. Estimates are computed by this site from actual file sizes plus runtime overhead (method<\/a>); long contexts need more.<\/p>\n

25 models listed.<\/em><\/p>\n

Everything that fits your GPU: 8GB<\/a> \/ 12GB<\/a> \/ 16GB<\/strong> \/ 24GB<\/a><\/p>\n

Text Generation (Chat and Code)<\/h2>\n
\n

\u2192 Scroll horizontally to see all columns<\/p>\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n
Model<\/th>\nParameters<\/th>\nBest build for 16GB<\/th>\nEst. memory<\/th>\nSmallest tier<\/th>\nRuns on<\/th>\nArticle<\/th>\n<\/tr>\n<\/thead>\n
nex-agi\/Nex-N2.5-mini<\/td>\n35.1B<\/td>\nQ2_K (12.9GB)<\/td>\n15.5GB<\/td>\n16GB<\/td>\nOllama \/ LM Studio, vLLM, MLX<\/td>\nRead<\/a><\/td>\n<\/tr>\n
prism-ml\/Ternary-Bonsai-2-27B-gguf<\/td>\n27.8B<\/td>\nQ2_0 (6.7GB)<\/td>\n8.1GB<\/td>\n8GB<\/td>\nOllama \/ LM Studio, MLX<\/td>\nRead<\/a><\/td>\n<\/tr>\n
bartowski\/Gryphe_Pantheon-Reasoning-26B-A4B-1.1-V2-GGUF<\/td>\n26.5B<\/td>\nIQ3_XS (12.7GB)<\/td>\n15.2GB<\/td>\n12GB<\/td>\nOllama \/ LM Studio<\/td>\nRead<\/a><\/td>\n<\/tr>\n
togethercomputer\/Tev1-4B-experimental<\/td>\n4.7B<\/td>\nBF16 (8.7GB)<\/td>\n10.4GB<\/td>\n12GB<\/td>\nvLLM<\/td>\nRead<\/a><\/td>\n<\/tr>\n
tencent\/Simple-Attention-Sparsification<\/td>\n4.0B<\/td>\nBF16 (7.5GB)<\/td>\n9.0GB<\/td>\n12GB<\/td>\n–<\/td>\nRead<\/a><\/td>\n<\/tr>\n
openbmb\/MiniCPM5-2B<\/td>\n2.5B<\/td>\nF16 (4.7GB)<\/td>\n5.6GB<\/td>\n4GB<\/td>\nOllama \/ LM Studio, vLLM, MLX<\/td>\nRead<\/a><\/td>\n<\/tr>\n
harshatheg\/Qwen-2.5-1B-RLCD<\/td>\n1.5B<\/td>\nBF16 (2.9GB)<\/td>\n3.5GB<\/td>\n4GB<\/td>\nMLX<\/td>\nRead<\/a><\/td>\n<\/tr>\n
pfnet\/plamo-3-610m-fin-instruct<\/td>\n890M<\/td>\nBF16 (1.7GB)<\/td>\n2.0GB<\/td>\n4GB<\/td>\nvLLM<\/td>\nRead<\/a><\/td>\n<\/tr>\n
togethercomputer\/Tev1-0.8B-experimental<\/td>\n873M<\/td>\nF32 (1.6GB)<\/td>\n2.0GB<\/td>\n4GB<\/td>\nvLLM<\/td>\nRead<\/a><\/td>\n<\/tr>\n
Cactus-Compute\/needle3<\/td>\n–<\/td>\nOriginal precision (0.2GB)<\/td>\n0.3GB<\/td>\n4GB<\/td>\n–<\/td>\nRead<\/a><\/td>\n<\/tr>\n<\/tbody>\n<\/table>\n<\/div>\n

Vision-Language and Multimodal Models<\/h2>\n
\n

\u2192 Scroll horizontally to see all columns<\/p>\n\n\n\n\n\n\n\n\n\n\n\n
Model<\/th>\nParameters<\/th>\nBest build for 16GB<\/th>\nEst. memory<\/th>\nSmallest tier<\/th>\nRuns on<\/th>\nArticle<\/th>\n<\/tr>\n<\/thead>\n
bartowski\/vectionlabs_Salience-27B-R6-GGUF<\/td>\n27.8B<\/td>\nQ3_K_L (13.2GB)<\/td>\n15.8GB<\/td>\n12GB<\/td>\nOllama \/ LM Studio<\/td>\nRead<\/a><\/td>\n<\/tr>\n
DavidAU\/Qwen3.8-27B-TWIN-TURBO-Fable-Cold-Fusion-709-L-Uncensored-NM-DAU-NEO-MTP-GGUF<\/td>\n27.8B<\/td>\nIQ3_M (11.7GB)<\/td>\n14.1GB<\/td>\n12GB<\/td>\nOllama \/ LM Studio<\/td>\nRead<\/a><\/td>\n<\/tr>\n
agentionai\/Signal-3.8-27B-GGUF<\/td>\n27.8B<\/td>\nIQ4_XS (13.3GB)<\/td>\n15.9GB<\/td>\n16GB<\/td>\nOllama \/ LM Studio<\/td>\nRead<\/a><\/td>\n<\/tr>\n
Jackrong\/Qwopus3.8-27B-Flash-GGUF<\/td>\n27.8B<\/td>\nQ3_K_M (12.6GB)<\/td>\n15.1GB<\/td>\n16GB<\/td>\nOllama \/ LM Studio<\/td>\nRead (Japanese)<\/a><\/td>\n<\/tr>\n
bartowski\/TheDrummer_Orion-26B-A4B-v1.1-GGUF<\/td>\n25.8B<\/td>\nIQ3_M (13.4GB)<\/td>\n16.0GB<\/td>\n12GB<\/td>\nOllama \/ LM Studio<\/td>\nRead<\/a><\/td>\n<\/tr>\n
ggml-org\/MiMo-V2.6-Distill-Qwen-9B-GGUF<\/td>\n9.4B<\/td>\nQ8_0 (8.9GB)<\/td>\n10.6GB<\/td>\n12GB<\/td>\nOllama \/ LM Studio<\/td>\nRead<\/a><\/td>\n<\/tr>\n
LiquidAI\/LFM2.5-VL-3B-DSpark<\/td>\n279M<\/td>\nF16 (0.5GB)<\/td>\n0.6GB<\/td>\n4GB<\/td>\n–<\/td>\nRead<\/a><\/td>\n<\/tr>\n<\/tbody>\n<\/table>\n<\/div>\n

Image Generation and Editing<\/h2>\n
\n

\u2192 Scroll horizontally to see all columns<\/p>\n\n\n\n\n\n
Model<\/th>\nParameters<\/th>\nBest build for 16GB<\/th>\nEst. memory<\/th>\nSmallest tier<\/th>\nRuns on<\/th>\nArticle<\/th>\n<\/tr>\n<\/thead>\n
abenzerps\/Qwen-Image-2.1-Uncensored-GGUF<\/td>\n7.1B<\/td>\nBF16 (13.3GB)<\/td>\n15.9GB<\/td>\n8GB<\/td>\n–<\/td>\nRead<\/a><\/td>\n<\/tr>\n<\/tbody>\n<\/table>\n<\/div>\n

Video Generation<\/h2>\n
\n

\u2192 Scroll horizontally to see all columns<\/p>\n\n\n\n\n\n
Model<\/th>\nParameters<\/th>\nBest build for 16GB<\/th>\nEst. memory<\/th>\nSmallest tier<\/th>\nRuns on<\/th>\nArticle<\/th>\n<\/tr>\n<\/thead>\n
WarmBloodAban\/Minimax-h3_Singularity<\/td>\n33.1B<\/td>\nW4A8 (11.0GB)<\/td>\n13.2GB<\/td>\n16GB<\/td>\n–<\/td>\nRead (Japanese)<\/a><\/td>\n<\/tr>\n<\/tbody>\n<\/table>\n<\/div>\n

Audio: Speech Synthesis, Music and Speech Recognition<\/h2>\n
\n

\u2192 Scroll horizontally to see all columns<\/p>\n\n\n\n\n\n\n\n
Model<\/th>\nParameters<\/th>\nBest build for 16GB<\/th>\nEst. memory<\/th>\nSmallest tier<\/th>\nRuns on<\/th>\nArticle<\/th>\n<\/tr>\n<\/thead>\n
Edge0\/Audio8-ASR-Infinite<\/td>\n4.1B<\/td>\nBF16 (7.6GB)<\/td>\n9.1GB<\/td>\n12GB<\/td>\n–<\/td>\nRead<\/a><\/td>\n<\/tr>\n
m-a-p\/YuE2-3B<\/td>\n3.6B<\/td>\nBF16 (6.8GB)<\/td>\n8.1GB<\/td>\n12GB<\/td>\n–<\/td>\nRead<\/a><\/td>\n<\/tr>\n
phasefield-audio\/Irodori-TTS-v4.1-Anime<\/td>\n766M<\/td>\nF32 (2.9GB)<\/td>\n3.4GB<\/td>\n4GB<\/td>\n–<\/td>\nRead<\/a><\/td>\n<\/tr>\n<\/tbody>\n<\/table>\n<\/div>\n

Task not declared<\/h2>\n
\n

\u2192 Scroll horizontally to see all columns<\/p>\n\n\n\n\n\n\n\n
Model<\/th>\nParameters<\/th>\nBest build for 16GB<\/th>\nEst. memory<\/th>\nSmallest tier<\/th>\nRuns on<\/th>\nArticle<\/th>\n<\/tr>\n<\/thead>\n
tencent\/WeVisDoc-4B<\/td>\n4.4B<\/td>\nF16 (8.2GB)<\/td>\n9.9GB<\/td>\n4GB<\/td>\n–<\/td>\nRead<\/a><\/td>\n<\/tr>\n
Comfy-Org\/YuE2<\/td>\n3.6B<\/td>\nBF16 (6.8GB)<\/td>\n8.1GB<\/td>\n12GB<\/td>\n–<\/td>\nRead<\/a><\/td>\n<\/tr>\n
Efficient-Large-Model\/H3-to-LTX-Latent-Adapter<\/td>\n195M<\/td>\nBF16 (0.4GB)<\/td>\n0.4GB<\/td>\n4GB<\/td>\n–<\/td>\nRead<\/a><\/td>\n<\/tr>\n<\/tbody>\n<\/table>\n<\/div>\n","protected":false},"excerpt":{"rendered":"

Every model covered by Local Model Watch that fits in 16GB of GPU memory, with the best build that fits. The “Best build for 16GB” column is the largest (highest-quality) quantization or precision whose estimated memory stays within 16GB \u2014 often better than the smallest build listed in the VRAM quick reference. Estimates are computed […]<\/p>\n","protected":false},"author":1,"featured_media":0,"parent":0,"menu_order":0,"comment_status":"closed","ping_status":"closed","template":"","meta":{"footnotes":""},"class_list":["post-3547","page","type-page","status-publish","hentry"],"lang":"en","translations":{"en":3547,"ja":3546},"_links":{"self":[{"href":"https:\/\/localmodelwatch.tsuchitsuchi.com\/en\/wp-json\/wp\/v2\/pages\/3547","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/localmodelwatch.tsuchitsuchi.com\/en\/wp-json\/wp\/v2\/pages"}],"about":[{"href":"https:\/\/localmodelwatch.tsuchitsuchi.com\/en\/wp-json\/wp\/v2\/types\/page"}],"author":[{"embeddable":true,"href":"https:\/\/localmodelwatch.tsuchitsuchi.com\/en\/wp-json\/wp\/v2\/users\/1"}],"replies":[{"embeddable":true,"href":"https:\/\/localmodelwatch.tsuchitsuchi.com\/en\/wp-json\/wp\/v2\/comments?post=3547"}],"version-history":[{"count":2,"href":"https:\/\/localmodelwatch.tsuchitsuchi.com\/en\/wp-json\/wp\/v2\/pages\/3547\/revisions"}],"predecessor-version":[{"id":4456,"href":"https:\/\/localmodelwatch.tsuchitsuchi.com\/en\/wp-json\/wp\/v2\/pages\/3547\/revisions\/4456"}],"wp:attachment":[{"href":"https:\/\/localmodelwatch.tsuchitsuchi.com\/en\/wp-json\/wp\/v2\/media?parent=3547"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}