BGE Reranker
bge-reranker-v2-m3
Weight size basis: 0.57 B
License identifier (see source): apache-2.0
Use: Search reranking
Unknown
Weight estimate GiB: 0.3 · 4-bit / 4096
Architecture/format needs review
AQI AI / HARDWARE
Find models for my hardware
Capacity estimate only; no automatic download or connection to your computer.
A growing local model snapshot, separated by chat, coding, retrieval, reranking and speech tasks. It spans generations, not the entire market or every latest release. Quality and speed are not uniformly benchmarked here.
Text embedding · Search reranking · Speech to text
Embedding models produce retrieval vectors, rerankers score candidate passages, and speech models transcribe audio. These tasks are not comparable by chat quality. Batch, activation and audio memory are not modeled; specialist capacity stays unverified instead of using the chat-cache formula.
Find models for my hardware
Calculated only in this page; no upload, automatic detection or model-name storage. CPU/GPU names do not predict speed; verify drivers and runtime support separately.
Capacity estimate only; no automatic download or connection to your computer.
Capacity scenario, not a benchmark
File sizes and hashes come from author repository listings; weights were not downloaded. File size is not total runtime memory. A verified matching-precision file takes precedence over the generic weight estimate, before cache and runtime margins.
Capacity scenario, not a benchmark: public weight tensor elements × bits ÷ 8 plus a 20% quantization/mixed-precision margin; F16 KV from layers, KV heads and context, plus 1.5 GiB runtime margin. Reserve at least 4 GiB or 25% RAM, and 2 GiB dedicated VRAM. One model, one text sequence; not a measured GGUF size or run guarantee.
Hybrid architectures, special packing or missing configs require manual review. Image/audio overhead is excluded. MoE uses total weights, not active parameters. 4/8-bit scenarios do not establish a published quant or backend compatibility.
Filtered by capacity, not ranked by quality; verify runtime and actual speed.
Browse model catalog · 46
46
DeepSeek R1 Distill
DeepSeek-R1-Distill-Qwen-1.5B
Weight size basis: 1.78 B
License identifier (see source): mit + Apache-2.0 base
Use: Text chat · Reasoning
Documented/configured context limit: 131,072
Weight estimate GiB: 1.0 · 4-bit / 4096
DeepSeek R1 Distill
DeepSeek-R1-Distill-Qwen-7B
Weight size basis: 7.62 B
License identifier (see source): mit + Apache-2.0 base
Use: Text chat · Reasoning
Documented/configured context limit: 131,072
Weight estimate GiB: 4.3 · 4-bit / 4096
DeepSeek R1 Distill
DeepSeek-R1-Distill-Llama-8B
Weight size basis: 8.03 B
License identifier (see source): mit + Llama 3.1 base
Use: Text chat · Reasoning
Documented/configured context limit: 131,072
Weight estimate GiB: 4.5 · 4-bit / 4096
DeepSeek R1 Distill
DeepSeek-R1-Distill-Qwen-14B
Weight size basis: 14.77 B
License identifier (see source): mit + Apache-2.0 base
Use: Text chat · Reasoning
Documented/configured context limit: 131,072
Weight estimate GiB: 8.3 · 4-bit / 4096
DeepSeek R1 Distill
DeepSeek-R1-Distill-Qwen-32B
Weight size basis: 32.76 B
License identifier (see source): mit + Apache-2.0 base
Use: Text chat · Reasoning
Documented/configured context limit: 131,072
Weight estimate GiB: 18.3 · 4-bit / 4096
DeepSeek R1 Distill
DeepSeek-R1-Distill-Llama-70B
Weight size basis: 70.55 B
License identifier (see source): mit + Llama 3.3 base
Use: Text chat · Reasoning
Documented/configured context limit: 131,072
Weight estimate GiB: 39.4 · 4-bit / 4096
Gemma 3
gemma-3-1b-it
Weight size basis: 1.00 B
License identifier (see source): gemma
Use: Text chat
Unknown
Weight estimate GiB: 0.6 · 4-bit / 4096
Architecture/format needs review
Gemma 3
gemma-3-4b-it
Weight size basis: 4.30 B
License identifier (see source): gemma
Use: Text chat · Vision / multimodal
Unknown
Weight estimate GiB: 2.4 · 4-bit / 4096
Architecture/format needs review
Gemma 3
gemma-3-12b-it
Weight size basis: 12.19 B
License identifier (see source): gemma
Use: Text chat · Vision / multimodal
Unknown
Weight estimate GiB: 6.8 · 4-bit / 4096
Architecture/format needs review
Gemma 3
gemma-3-27b-it
Weight size basis: 27.43 B
License identifier (see source): gemma
Use: Text chat · Vision / multimodal
Unknown
Weight estimate GiB: 15.3 · 4-bit / 4096
Architecture/format needs review
Gemma 4
gemma-4-12B
Weight size basis: 11.96 B
License identifier (see source): apache-2.0
Use: Text chat · Vision / multimodal
Documented/configured context limit: 262,144
Weight estimate GiB: 6.7 · 4-bit / 4096
Architecture/format needs review
gpt-oss
gpt-oss-20b
Weight size basis: 20.91 B
License identifier (see source): apache-2.0
Use: Text chat · Reasoning
Documented/configured context limit: 131,072
Weight estimate GiB: 11.7 · 4-bit / 4096
Architecture/format needs review
gpt-oss
gpt-oss-120b
Weight size basis: 116.83 B
License identifier (see source): apache-2.0
Use: Text chat · Reasoning
Documented/configured context limit: 131,072
Weight estimate GiB: 65.3 · 4-bit / 4096
Architecture/format needs review
Llama
Llama-3.2-1B-Instruct
Weight size basis: 1.24 B
License identifier (see source): llama3.2
Use: Text chat
Unknown
Weight estimate GiB: 0.7 · 4-bit / 4096
Architecture/format needs review
Llama
Llama-3.2-3B-Instruct
Weight size basis: 3.21 B
License identifier (see source): llama3.2
Use: Text chat
Unknown
Weight estimate GiB: 1.8 · 4-bit / 4096
Architecture/format needs review
Llama
Llama-3.1-8B-Instruct
Weight size basis: 8.03 B
License identifier (see source): llama3.1
Use: Text chat
Unknown
Weight estimate GiB: 4.5 · 4-bit / 4096
Architecture/format needs review
Llama
Llama-3.3-70B-Instruct
Weight size basis: 70.55 B
License identifier (see source): llama3.3
Use: Text chat
Unknown
Weight estimate GiB: 39.4 · 4-bit / 4096
Architecture/format needs review
Ministral 3
Ministral-3-3B-Instruct-2512-BF16
Weight size basis: 4.25 B
License identifier (see source): apache-2.0
Use: Text chat
Documented/configured context limit: 262,144
Weight estimate GiB: 2.4 · 4-bit / 4096
Ministral 3
Ministral-3-8B-Instruct-2512-BF16
Weight size basis: 8.92 B
License identifier (see source): apache-2.0
Use: Text chat
Documented/configured context limit: 262,144
Weight estimate GiB: 5.0 · 4-bit / 4096
Ministral 3
Ministral-3-14B-Instruct-2512-BF16
Weight size basis: 13.95 B
License identifier (see source): apache-2.0
Use: Text chat
Documented/configured context limit: 262,144
Weight estimate GiB: 7.8 · 4-bit / 4096
Mistral Small
Mistral-Small-3.1-24B-Instruct-2503
Weight size basis: 24.01 B
License identifier (see source): apache-2.0
Use: Text chat
Documented/configured context limit: 131,072
Weight estimate GiB: 13.4 · 4-bit / 4096
Phi-4
Phi-4-mini-instruct
Weight size basis: 3.84 B
License identifier (see source): mit
Use: Text chat
Documented/configured context limit: 131,072
Weight estimate GiB: 2.1 · 4-bit / 4096
Phi-4
phi-4
Weight size basis: 14.66 B
License identifier (see source): mit
Use: Text chat
Documented/configured context limit: 16,384
Weight estimate GiB: 8.2 · 4-bit / 4096
Qwen3
Qwen3-0.6B
Weight size basis: 0.75 B
License identifier (see source): apache-2.0
Use: Text chat
Documented/configured context limit: 32,768
Weight estimate GiB: 0.4 · 4-bit / 4096
Qwen3
Qwen3-1.7B
Weight size basis: 2.03 B
License identifier (see source): apache-2.0
Use: Text chat
Documented/configured context limit: 32,768
Weight estimate GiB: 1.1 · 4-bit / 4096
Qwen3
Qwen3-4B
Weight size basis: 4.02 B
License identifier (see source): apache-2.0
Use: Text chat
Documented/configured context limit: 32,768
Weight estimate GiB: 2.3 · 4-bit / 4096
Qwen3
Qwen3-8B
Weight size basis: 8.19 B
License identifier (see source): apache-2.0
Use: Text chat
Documented/configured context limit: 32,768
Weight estimate GiB: 4.7 · 4-bit / 4096
Qwen3
Qwen3-14B
Weight size basis: 14.77 B
License identifier (see source): apache-2.0
Use: Text chat
Documented/configured context limit: 32,768
Weight estimate GiB: 8.3 · 4-bit / 4096
Qwen3
Qwen3-30B-A3B
Weight size basis: 30.53 B
License identifier (see source): apache-2.0
Use: Text chat
Documented/configured context limit: 32,768
Weight estimate GiB: 17.1 · 4-bit / 4096
Qwen3
Qwen3-32B
Weight size basis: 32.76 B
License identifier (see source): apache-2.0
Use: Text chat
Documented/configured context limit: 32,768
Weight estimate GiB: 18.3 · 4-bit / 4096
Qwen3 Coder
Qwen3-Coder-30B-A3B-Instruct
Weight size basis: 30.53 B
License identifier (see source): apache-2.0
Use: Text chat · Code focused
Documented/configured context limit: 262,144
Weight estimate GiB: 17.1 · 4-bit / 4096
Qwen3 Embedding
Qwen3-Embedding-0.6B
Weight size basis: 0.60 B
License identifier (see source): apache-2.0
Use: Text embedding
Documented/configured context limit: 32,768
Weight estimate GiB: 0.3 · 4-bit / 4096
Architecture/format needs review
Qwen3 Embedding
Qwen3-Embedding-4B
Weight size basis: 4.02 B
License identifier (see source): apache-2.0
Use: Text embedding
Documented/configured context limit: 32,768
Weight estimate GiB: 2.2 · 4-bit / 4096
Architecture/format needs review
Qwen3 Embedding
Qwen3-Embedding-8B
Weight size basis: 7.57 B
License identifier (see source): apache-2.0
Use: Text embedding
Documented/configured context limit: 32,768
Weight estimate GiB: 4.2 · 4-bit / 4096
Architecture/format needs review
Qwen3 Reranker
Qwen3-Reranker-0.6B
Weight size basis: 0.60 B
License identifier (see source): apache-2.0
Use: Search reranking
Documented/configured context limit: 32,768
Weight estimate GiB: 0.3 · 4-bit / 4096
Architecture/format needs review
Qwen3 Reranker
Qwen3-Reranker-4B
Weight size basis: 4.02 B
License identifier (see source): apache-2.0
Use: Search reranking
Documented/configured context limit: 32,768
Weight estimate GiB: 2.2 · 4-bit / 4096
Architecture/format needs review
Qwen3 Reranker
Qwen3-Reranker-8B
Weight size basis: 8.19 B
License identifier (see source): apache-2.0
Use: Search reranking
Documented/configured context limit: 32,768
Weight estimate GiB: 4.6 · 4-bit / 4096
Architecture/format needs review
Qwen3.5
Qwen3.5-4B
Weight size basis: 4.66 B
License identifier (see source): apache-2.0
Use: Text chat · Vision / multimodal
Documented/configured context limit: 262,144
Weight estimate GiB: 2.6 · 4-bit / 4096
Architecture/format needs review
Qwen3.5
Qwen3.5-9B
Weight size basis: 9.65 B
License identifier (see source): apache-2.0
Use: Text chat · Vision / multimodal
Documented/configured context limit: 262,144
Weight estimate GiB: 5.4 · 4-bit / 4096
Architecture/format needs review
Qwen3.5
Qwen3.5-27B
Weight size basis: 27.78 B
License identifier (see source): apache-2.0
Use: Text chat · Vision / multimodal
Documented/configured context limit: 262,144
Weight estimate GiB: 15.5 · 4-bit / 4096
Architecture/format needs review
Qwen3.8
Qwen3.8-27B
Weight size basis: 27.78 B
License identifier (see source): apache-2.0
Use: Text chat · Vision / multimodal
Documented/configured context limit: 262,144
Weight estimate GiB: 15.5 · 4-bit / 4096
Architecture/format needs review
SmolLM2
SmolLM2-135M-Instruct
Weight size basis: 0.13 B
License identifier (see source): apache-2.0
Use: Text chat
Documented/configured context limit: 8,192
Weight estimate GiB: 0.1 · 4-bit / 4096
SmolLM2
SmolLM2-360M-Instruct
Weight size basis: 0.36 B
License identifier (see source): apache-2.0
Use: Text chat
Documented/configured context limit: 8,192
Weight estimate GiB: 0.2 · 4-bit / 4096
SmolLM2
SmolLM2-1.7B-Instruct
Weight size basis: 1.71 B
License identifier (see source): apache-2.0
Use: Text chat
Documented/configured context limit: 8,192
Weight estimate GiB: 1.0 · 4-bit / 4096
Whisper
whisper-large-v3-turbo
Weight size basis: 0.81 B
License identifier (see source): mit
Use: Speech to text
Unknown
Weight estimate GiB: 0.5 · 4-bit / 4096
Architecture/format needs review
No matching models.