AI Model Hub
Browse AI models for multimodal decomposition and recomposition pipelines: plug any model into your extractors.
17,271 models available
Showing 97-120 of 17,271 models
Automatic Speech Recognition
jonatasgrosman/wav2vec2-large-xlsr-53-dutch
3.7M
17
transformersAutomatic Speech Recognition
jonatasgrosman/wav2vec2-large-xlsr-53-greek
3.3M
4
transformers......
5 / 720Choosing a model for multimodal retrieval
Which of these models does Mixpeek actually run?
Nine, and they are not the same thing as the catalog. The managed extractors run intfloat/multilingual-e5-large-instruct for text (1024-d), google/siglip-base-patch16-224 for images (768-d), google/vertex-multimodal (1408-d) and google/gemini-embedding-2 (3072-d) for unified multimodal, insightface ArcFace for faces (512-d), CLAP for audio fingerprints (512-d), facebook/dinov2-base for visual similarity and jinaai/jina-embeddings-v2-base-code for code (both 768-d, inside the web scraper). The rerank retriever stage runs BAAI/bge-reranker-v2-m3. Everything else in this catalog is documented here, not hosted here.
Can I run any model from this catalog on Mixpeek?
Not by naming it. No extractor takes a Hugging Face model id as a parameter, so there is no field to put one in. Three paths do work. Use a managed extractor and get the model it runs. Run the model on your own hardware and upsert the vectors through POST /v1/namespaces/{namespace_id}/documents/upsert, which stores them beside everything else. Or on Enterprise, upload the weights through POST /v1/namespaces/{namespace_id}/models, which accepts the huggingface format, and load them from a custom plugin.
Should I use a text embedding model or a multimodal one?
Ask whether a text query has to reach a non-text asset directly. If your video is searchable through its transcript and your images through their captions, a text model over that generated text is cheaper and usually more accurate, because retrieval quality on words is a solved problem and cross-modal alignment is not. If the query is 'find the shot that looks like this' or the visual content carries meaning no caption records, you need a shared space and a multimodal model. Most production systems run both indexes rather than choosing.
What does the embedding dimension cost to store?
A float32 vector is 4 bytes per dimension, so a million items costs 4 GB at 1024 dimensions, 3 GB at 768, and 12 GB at 3072. That is before any quantization and before payload. Models trained with Matryoshka representation learning, such as Gemini Embedding 2, let you truncate to a shorter prefix without re-encoding, so the width becomes an index-time decision rather than a model-selection one. Dimensions are fixed at namespace creation in Mixpeek, so changing a width later means re-indexing.
Why does the download count on a model page differ from HuggingFace?
It is the monthly figure from the HuggingFace API at the time of the last sync, not a live read, so it lags. The sync date is on each model page. Download count is a popularity signal and a poor quality signal: the most-downloaded model in a category is frequently an older checkpoint that a tutorial pinned years ago.
What's new in models
Full changelog- Sep 3, 2026Meta's Muse Spark joins the model registryThe model registry has a page for Muse Spark, Meta Superintelligence Labs' multimodal model: what it takes in, what it returns, its context window and licence, and how to route retriever results to it through an OpenAI-compatible call, alongside pipelines that pair it with an embedding model and a transcription model.
- Jun 22, 2026200+ model pages and new retrieval deep-divesExpanded the model catalog with enriched pages (architecture, benchmarks, Mixpeek SDK examples, and pipeline pairings) across embedding, vision-language, video, audio, OCR, and reranking models: plus new vendor-neutral guides on Matryoshka embeddings, reasoning rerankers, and retrieval feedback loops.
All models
Every model page in one place, 375 in total.
- ahmed-masry/ColMate-3B
- AIDC-AI/Ovis2.6-30B-A3B
- AILab-CVC/YOLO-World-L
- Alibaba-NLP/gme-Qwen2-VL-2B-Instruct
- Alibaba-NLP/gte-modernbert-base
- Alibaba-NLP/gte-reranker-modernbert-base
- Alibaba-NLP/GVE-3B
- allenai/Molmo2-8B
- allenai/olmOCR-2-7B-1025
- amazon/chronos-2
- amazon/chronos-bolt-base
- amazon/patchcore-resnet50
- apple/aimv2-3B-patch14-448
- apple/aimv2-large-patch14-native
- apple/AIMv2-large-patch14-native
- apple/DepthPro
- apple/FastVLM-0.5B
- apple/FastVLM-7B
- BAAI/bge-large-en-v1.5
- BAAI/bge-m3
- BAAI/bge-multilingual-gemma2
- BAAI/bge-reranker-v2-m3
- BAAI/bge-reranker-v2.5-gemma2-lightweight
- BAAI/BGE-VL-base
- BAAI/BGE-VL-v1.5-zs
- BAAI/EVA02-CLIP-L-14-336
- baidu/Qianfan-OCR
- baidu/Unlimited-OCR
- BidirLM/BidirLM-Omni-2.5B-Embedding
- briaai/RMBG-1.4
- briaai/RMBG-2.0
- bytedance-research/Lance
- bytedance-research/Vidi-7B
- ByteDance/Dolphin-v2
- CIDAS/clipseg-rd64-refined
- codefuse-ai/F2LLM-v2-14B
- Cognitive-Lab/ColNetraEmbed
- CohereLabs/cohere-transcribe-03-2026
- CohereLabs/cohere-transcribe-arabic-07-2026
- CohereLabs/command-a-plus-05-2026-bf16
- ContextualAI/ctxl-rerank-v2-instruct-multilingual-1b
- cross-encoder/ettin-reranker-150m-v1
- cross-encoder/ettin-reranker-17m-v1
- cross-encoder/ettin-reranker-1b-v1
- cross-encoder/ettin-reranker-32m-v1
- cross-encoder/ettin-reranker-400m-v1
- cross-encoder/ettin-reranker-68m-v1
- DAMO-NLP-SG/VideoLLaMA3-7B
- Datadog/Toto-2.0-2.5B
- datalab-to/chandra-ocr-2
- deepinsight/insightface-retinaface-r50
- deepseek-ai/DeepSeek-OCR-2
- deepseek-ai/DeepSeek-V4-Flash-Vision-Exp
- deepseek-ai/deepseek-vl2-small
- depth-anything/DA3-LARGE-1.1
- depth-anything/DA3-SMALL
- depth-anything/DA3NESTED-GIANT-LARGE-1.1
- depth-anything/Depth-Anything-V2-Large
- distil-whisper/distil-large-v3
- docling-project/docling-layout-heron
- docling-project/docling-models
- docling-project/SmolDocling-256M-preview
- facebook/detr-resnet-50
- facebook/dinov2-base
- facebook/dinov2-large
- facebook/dinov3-convnext-large-pretrain-lvd1689m
- facebook/dinov3-large
- facebook/dinov3-vit7b16-pretrain-lvd1689m
- facebook/dinov3-vitb16-pretrain-lvd1689m
- facebook/dinov3-vitl16-pretrain-lvd1689m
- facebook/dinov3-vits16-pretrain-lvd1689m
- facebook/EdgeTAM
- facebook/encodec_24khz
- facebook/encodec_48khz
- facebook/faiss
- facebook/hubert-large-ls960-ft
- facebook/metaclip-2-worldwide-huge-quickgelu
- facebook/mms-1b-all
- facebook/pe-av-large
- facebook/PE-Core-G14-448
- facebook/PE-Spatial-G14-448
- facebook/Perception-LM-3B
- facebook/sam-vit-huge
- facebook/sam2.1-hiera-base-plus
- facebook/sam2.1-hiera-large
- facebook/sam3
- facebook/sam3.1
- facebook/seamless-m4t-v2-large
- facebook/VGGT-1B
- facebook/vjepa2-vitg-fpc64-256
- facebook/vjepa2-vitl-fpc64-256
- facebook/wav2vec2-large-960h
- fal/AuraFace-v1
- fastino/gliner2-base-v1
- FireRedTeam/FireRed-OCR
- FireRedTeam/ReMatch-3B
- foduucom/table-detection-and-extraction
- FunAudioLLM/SenseVoiceSmall
- fushh7/ObjEmbed-2B
- google/embeddinggemma-300m
- google/gemini-embedding-2
- google/gemma-4-12B-it
- google/gemma-4-26B-A4B-it
- google/gemma-4-31B-it
- google/gemma-4-E2B-it
- google/gemma-4-E4B-it
- google/medgemma-27b-it
- google/owlv2-large-patch14-ensemble
- google/owlvit-large-patch14
- google/paligemma2-3b-mix-448
- google/pix2struct-base
- google/scann
- google/siglip-base-patch16-224
- google/siglip-so400m-patch14-384
- google/siglip2-base-patch16-224
- google/siglip2-giant-opt-patch16-384
- google/siglip2-so400m-patch14-384
- google/siglip2-so400m-patch16-384
- google/siglip2-so400m-patch16-naflex
- google/timesfm-2.0-500m-pytorch
- google/timesfm-2.5-200m-transformers
- google/tipsv2-so400m14
- google/videoprism-base-f16r288
- google/videoprism-large-f8r288
- Haon-Chen/e5-omni-7B
- Hcompany/Holo-3.1-4B
- HuggingFaceTB/SmolVLM2-2.2B-Instruct
- hustvl/yolos-tiny
- ibm-granite/granite-4.0-1b-speech
- ibm-granite/granite-4.0-3b-vision
- ibm-granite/granite-docling-258M
- ibm-granite/granite-embedding-311m-multilingual-r2
- ibm-granite/granite-embedding-97m-multilingual-r2
- ibm-granite/granite-embedding-english-r2
- ibm-granite/granite-embedding-reranker-english-r2
- ibm-granite/granite-embedding-small-english-r2
- ibm-granite/granite-speech-4.1-2b
- ibm-granite/granite-speech-4.1-2b-plus
- ibm-granite/granite-vision-4.1-4b
- IDEA-Research/grounding-dino-base
- Idiap/EdgeFace-S-GAMMA
- immich-app/buffalo_l
- Intel/dpt-large
- intfloat/multilingual-e5-large-instruct
- intfloat/multilingual-e5-small
- iSEE-Laboratory/llmdet_large
- jinaai/jina-clip-v2
- jinaai/jina-code-embeddings-1.5b
- jinaai/jina-embeddings-v2-base-code
- jinaai/jina-embeddings-v4
- jinaai/jina-embeddings-v5-omni-nano
- jinaai/jina-embeddings-v5-omni-nano-retrieval
- jinaai/jina-embeddings-v5-omni-small
- jinaai/jina-embeddings-v5-omni-small-retrieval
- jinaai/jina-embeddings-v5-text-nano
- jinaai/jina-embeddings-v5-text-small
- jinaai/jina-reranker-m0
- jinaai/jina-reranker-v3
- Kingsoft-LLM/QZhou-Embedding
- Kwai-Keye/Keye-VL-2.0-30B-A3B
- Kwai-Keye/Keye-VL-8B-Preview
- kyutai/mimi
- kyutai/stt-2.6b-en
- laion/clap-htsat-fused
- laion/clap-htsat-unfused
- laion/CLIP-ViT-bigG-14-laion2B-39B-b160k
- laion/CLIP-ViT-H-14-laion2B-s32B-b79K
- laion/larger_clap_general
- laion/larger_clap_music
- laion/larger_clap_music_and_speech
- LCO-Embedding/LCO-Embedding-Omni-3B
- LCO-Embedding/LCO-Embedding-Omni-7B
- LGAI-EXAONE/EXAONE-4.5-33B
- lightonai/Agent-ModernColBERT
- lightonai/GTE-ModernColBERT-v1
- lightonai/LateOn-Code
- lightonai/LightOnOCR-2-1B
- lightonai/Reason-ModernColBERT
- LiquidAI/LFM2.5-ColBERT-350M
- lpiccinelli/unidepth-v2-vitl14
- m-a-p/MERT-v1-330M
- m-a-p/MERT-v1-95M
- Maple728/TimeMoE-50M
- Marqo/marqo-fashionSigLIP
- meta/muse-spark
- microsoft/BEATs
- microsoft/codebert-base
- microsoft/dit-large
- microsoft/Fara-7B
- microsoft/Florence-2-large
- microsoft/harrier-oss-v1-0.6b
- microsoft/layoutlmv3-base
- microsoft/Mage-VL
- microsoft/OmniParser-v2.0
- microsoft/Phi-4-multimodal-instruct
- microsoft/Phi-4-reasoning-vision-15B
- microsoft/table-transformer-detection
- microsoft/table-transformer-structure-recognition-v1.1-all
- microsoft/table-transformer-structure-recognition-v1.1-pub
- microsoft/trocr-large-handwritten
- microsoft/trocr-large-printed
- microsoft/VibeVoice-ASR
- microsoft/VibeVoice-ASR-HF
- microsoft/wavlm-large
- minchul/cvlface_adaface_ir101_webface12m
- minchul/cvlface_adaface_ir50_ms1mv2
- minchul/cvlface_adaface_vit_base_kprpe_webface12m
- minchul/cvlface_adaface_vit_base_webface4m
- MiniMaxAI/MiniMax-M3
- mispeech/dasheng-1.2B
- mistralai/Voxtral-Mini-3B-2507
- mistralai/Voxtral-Mini-4B-Realtime-2602
- MIT/ast-finetuned-audioset-10-10-0.4593
- mixedbread-ai/mxbai-colbert-large-v1
- mixedbread-ai/mxbai-rerank-large-v2
- ModernVBERT/ColModernVBERT
- moondream/moondream3-preview
- moonshotai/Kimi-K2.6
- moonshotai/Kimi-VL-A3B-Thinking-2506
- moonshotai/MoonViT-SO-400M
- mtg-upf/discogs-maest-30s-pw-129e
- nanonets/Nanonets-OCR2-3B
- naver-clova-ix/donut-base
- NCSOFT/GME-VARCO-VISION-Embedding
- NemoStation/Marlin-2B
- netflix/void-model
- nomic-ai/CodeRankEmbed
- nomic-ai/colnomic-embed-multimodal-7b
- nomic-ai/modernbert-embed-base
- nomic-ai/modernbert-embed-large
- nomic-ai/nomic-embed-code
- nomic-ai/nomic-embed-multimodal-3b
- nomic-ai/nomic-embed-multimodal-7b
- nomic-ai/nomic-embed-text-v2-moe
- nomic-ai/nomic-embed-vision-v1.5
- numind/NuExtract3
- nvidia/4D-RGPT-8B
- nvidia/C-RADIOv4-H
- nvidia/C-RADIOv4-SO400M
- nvidia/canary-1b-v2
- nvidia/canary-qwen-2.5b
- nvidia/Cosmos-Embed1-448p-anomaly-detection
- nvidia/Cosmos-Reason2-2B
- nvidia/Cosmos3-Nano
- nvidia/diar_sortformer_4spk-v1
- nvidia/diar_streaming_sortformer_4spk-v2
- nvidia/Eagle2.5-8B
- nvidia/llama-embed-nemotron-8b
- nvidia/llama-nemotron-embed-vl-1b-v2
- nvidia/llama-nemotron-rerank-1b-v2
- nvidia/llama-nemotron-rerank-vl-1b-v2
- nvidia/LocateAnything-3B
- nvidia/Nemotron-3-Embed-1B-BF16
- nvidia/Nemotron-3-Embed-8B-BF16
- nvidia/Nemotron-3-Nano-Omni-30B-A3B-Reasoning-BF16
- nvidia/nemotron-3.5-asr-streaming-0.6b
- nvidia/nemotron-colembed-vl-8b-v2
- nvidia/nemotron-ocr-v2
- nvidia/nemotron-page-elements-v3
- nvidia/nemotron-speech-streaming-en-0.6b
- nvidia/NV-Embed-v2
- nvidia/NVIDIA-Nemotron-Parse-v1.1
- nvidia/omni-embed-nemotron-3b
- nvidia/parakeet-ctc-1.1b
- nvidia/parakeet-rnnt-1.1b
- nvidia/parakeet-tdt-0.6b-v3
- nvidia/parakeet-tdt-1.1b
- nvidia/speakerverification_en_titanet_large
- omlab/omdet-turbo-swin-tiny-hf
- omni-research/Tarsier2-7b-0115
- openai/clip-vit-base-patch32
- openai/clip-vit-large-patch14
- openai/whisper-large-v3
- openai/whisper-large-v3-turbo
- openbmb/MiniCPM-o-4_5
- openbmb/MiniCPM-V-4_5
- openbmb/MiniCPM-V-4.6
- opendatalab/MinerU2.5-Pro-2604-1.2B
- OpenGVLab/InternVideo2-Stage1-1B-224p-K700
- OpenGVLab/InternVL3_5-8B
- OpenGVLab/InternVL3-78B
- OpenGVLab/InternVL3-8B
- openmmlab-community/mm_grounding_dino_large_all
- PaddlePaddle/paddleocr
- PaddlePaddle/PaddleOCR-VL-1.5
- PaddlePaddle/PaddleOCR-VL-1.6
- PaddlePaddle/PP-DocLayoutV3
- ParaVT/ParaVT-8B
- PekingU/rtdetr_v2_r101vd
- perplexity-ai/pplx-embed-context-v1-4b
- perplexity-ai/pplx-embed-v1-0.6b
- perplexity-ai/pplx-embed-v1-late-0.6b
- poloclub/UniTable
- pyannote/speaker-diarization-3.1
- pyannote/speaker-diarization-community-1
- pyannote/wespeaker-voxceleb-resnet34-LM
- Qwen/Qwen3-ASR-0.6B
- Qwen/Qwen3-ASR-0.6B-hf
- Qwen/Qwen3-ASR-1.7B
- Qwen/Qwen3-Embedding-0.6B
- Qwen/Qwen3-Embedding-4B
- Qwen/Qwen3-Embedding-8B
- Qwen/Qwen3-ForcedAligner-0.6B
- Qwen/Qwen3-Reranker-0.6B
- Qwen/Qwen3-Reranker-4B
- Qwen/Qwen3-Reranker-8B
- Qwen/Qwen3-VL-30B-A3B-Instruct
- Qwen/Qwen3-VL-4B-Instruct
- Qwen/Qwen3-VL-8B-Instruct
- Qwen/Qwen3-VL-Embedding-2B
- Qwen/Qwen3-VL-Embedding-8B
- Qwen/Qwen3-VL-Reranker-2B
- Qwen/Qwen3-VL-Reranker-8B
- Qwen/Qwen3.6-27B
- Qwen/Qwen3.6-35B-A3B
- Qwen/Qwen3.8-27B
- rednote-hilab/dots.ocr
- reducto/RolmOCR
- roboflow/rf-detr-base
- Roboflow/rf-detr-base
- Roboflow/rf-detr-large
- Roboflow/rf-detr-medium
- Roboflow/rf-detr-seg-large
- Salesforce/blip2-opt-2.7b
- Salesforce/codet5p-110m-embedding
- Salesforce/moirai-2.0-R-small
- Salesforce/SFR-Embedding-Code-2B_R
- Salesforce/SFR-Embedding-Code-400M_R
- sensenova/SenseNova-U1-8B-MoT
- sentence-transformers/all-MiniLM-L6-v2
- sentence-transformers/LaBSE
- sentence-transformers/paraphrase-multilingual-MiniLM-L12-v2
- Snowflake/snowflake-arctic-embed-l-v2.0
- Snowflake/snowflake-arctic-embed-m-v2.0
- speechbrain/spkrec-ecapa-voxceleb
- stepfun-ai/GOT-OCR-2.0-hf
- stepfun-ai/Step-3.7-Flash
- Systran/faster-whisper-large-v3
- tencent/HunyuanOCR
- tencent/WeMM-Embedding-2B
- tencent/WeMM-Embedding-4B
- tencent/WeMM-Embedding-9B
- TencentARC/TimeLens-8B
- thebajajra/RexReranker-base
- thebajajra/RexReranker-large
- tiiuae/Falcon-OCR
- timesler/facenet-pytorch
- TomoroAI/tomoro-colqwen3-embed-4b
- tsinghua-ee/WAVE-7B
- ultralytics/yolo11n
- ultralytics/yolo26n
- ultralytics/yolov8n
- usefulsensors/moonshine-streaming-medium
- ustc-community/dfine-xlarge-coco
- vidore/colpali-v1.2
- vidore/colpali-v1.3
- vidore/colqwen-omni-v0.1
- vidore/colqwen2.5-v0.2
- Vision-CAIR/Tempo-6B
- VLM2Vec/VLM2Vec-V2.0
- voyageai/voyage-4-nano
- webAI-Official/webAI-ColVec1-4b
- Wespeaker/wespeaker-voxceleb-resnet293-LM
- WorldSeek-AI/WorldSeek-Omni-2B-Preview
- XiaomiMiMo/MiMo-V2.5-ASR
- yanziang/InternVideo3-8B-Instruct
- yyfz233/Pi3
- zai-org/GLM-4.5V
- zai-org/GLM-4.6V
- zai-org/GLM-5.3-Flash
- zai-org/GLM-OCR
- zeroentropy/zembed-1-embedding
- zeroentropy/zerank-2-reranker
- ZhengPeng7/BiRefNet
- ZhengPeng7/BiRefNet_HR