Which of these models does Mixpeek run today?

15 are served on Mixpeek as Ray Serve apps that start at zero replicas and wake on first use (Multilingual E5 Large Instruct, all-MiniLM-L6-v2, SigLIP Base patch16 224, InsightFace buffalo_l (SCRFD-10GF + ArcFace ResNet50 WebFace600K), InsightFace buffalo_sc (SCRFD-500MF + MobileFaceNet WebFace600K) and others). 13 are provider APIs Mixpeek calls by name. 34 can run as a custom plugin and do not run yet. 29 are blocked today; each page says why.

How is the default order decided?

By usage, the namespaces and collections that use each model, counted on 2026-11-06T13:42:25Z, then by how many templates use it, whether it runs today and whether we have our own evaluation.

What does unmeasured mean on a model?

We have not measured that model's cost or quality on our own image and data yet. Publisher benchmarks are linked from the model's sources and are not copied into the measured fields.

Can an agent read this catalog?

Yes. /models.json holds every record and /models/ .json holds one. They are the same records this page renders, generated from the catalog the Mixpeek server is built from.

How do I use a model in a namespace?

Open the model and deploy a template that already uses it, or add it to a new or existing namespace from the Designer in Studio. Every served model starts at zero replicas, so an idle namespace costs nothing for inference.

Models

Every model Mixpeek runs, calls or can load: what it is good and bad at, its inputs and outputs, how it runs with scale to zero, its license, and the templates that use it. Ordered by use. Filter by anything on the record; the address bar keeps the filter so you can share it.

91 models

Templates models.json

Recommended by use case

Multilingual E5 Large Instruct

Multilingual E5 Large Instruct

ColQwen2 v1.0 (merged)

BGE Reranker v2 M3

Falconsai NSFW image detection (ViT)

Whisper large-v3-turbo

SigLIP 2 so400m patch14 (224/384)

Gemini 3.5 Flash-Lite

Gemini 3.5 Flash-Lite

Gemini 3.5 Flash-Lite

D-FINE L (COCO)

SAM 2.1 Hiera Tiny (small, base-plus)

/

Task: Detection

12 of 91 models

Sort

Hardware, cost and evidence come from each record. "Unmeasured" means we have not measured it on our own image yet.

Runs as a custom plugin
D-FINE L (COCO)
Detection · ustc-community/dfine-large-coco
Real-time DETR with fine-grained box refinement; among the highest COCO AP per latency of the permissive detectors.
- In/out Image to Boxes, Labels, Scores
License
Apache-2.0
Hardware
Runs on CPU
Recommended
Yes
View modelJSON
Runs as a custom plugin
RF-DETR Base
Detection · Roboflow/rf-detr-base
DINOv2-backbone real-time detector that fine-tunes well on small custom datasets (RF100-VL); Nano to Large are Apache-2.0.
- In/out Image to Boxes, Labels, Scores
License
Apache-2.0
Hardware
Runs on CPU
View modelJSON
Runs as a custom plugin
RT-DETR R50vd
Detection · PekingU/rtdetr_r50vd
Real-time end-to-end detection with no NMS step; Apache-2.0, a drop-in for YOLO detect.
- In/out Image to Boxes, Labels, Scores
License
Apache-2.0
Hardware
Runs on CPU
View modelJSON
Runs as a custom plugin
RT-DETRv2 R50vd
Detection · PekingU/rtdetr_v2_r50vd
RT-DETR with a better training recipe (bag of freebies); same cost, higher COCO AP.
- In/out Image to Boxes, Labels, Scores
License
Apache-2.0
Hardware
Runs on CPU
View modelJSON
Runs as a custom plugin
YOLOX-S
Detection · yolox_s (Megvii release asset)
Mature Apache-2.0 YOLO-style detector with official ONNX exports; easy CPU deployment.
- In/out Image to Boxes, Labels, Scores
License
Apache-2.0
Hardware
Runs on CPU
View modelJSON
Not runnable today
Ultralytics YOLO11 detect
Detection · Ultralytics/YOLO11 (yolo11n.pt)
Fast detect for the fixed COCO label set; n size runs in real time on CPU.
- In/out Image to Boxes, Labels, Scores
License
AGPL-3.0
Hardware
Runs on CPU
View modelJSON
Not runnable today
Ultralytics YOLO11 pose
Detection · Ultralytics/YOLO11 (yolo11n-pose.pt)
Fast pose for the fixed COCO label set; n size runs in real time on CPU.
- In/out Image to Boxes, Keypoints, Scores
License
AGPL-3.0
Hardware
Runs on CPU
View modelJSON
Not runnable today
Ultralytics YOLO26 detect
Detection · Ultralytics/YOLO26 (yolo26n.pt)
Fast detect for the fixed COCO label set; n size runs in real time on CPU.
- In/out Image to Boxes, Labels, Scores
License
AGPL-3.0
Hardware
Runs on CPU
View modelJSON
Not runnable today
Ultralytics YOLO26 pose
Detection · Ultralytics/YOLO26 (yolo26n-pose.pt)
Fast pose for the fixed COCO label set; n size runs in real time on CPU.
- In/out Image to Boxes, Keypoints, Scores
License
AGPL-3.0
Hardware
Runs on CPU
View modelJSON
Not runnable today
Ultralytics YOLOv8 detect
Detection · Ultralytics/YOLOv8 (yolov8n.pt)
Fast detect for the fixed COCO label set; n size runs in real time on CPU.
- In/out Image to Boxes, Labels, Scores
License
AGPL-3.0
Hardware
Runs on CPU
View modelJSON
Not runnable today
Ultralytics YOLOv8 pose
Detection · Ultralytics/YOLOv8 (yolov8n-pose.pt)
Fast pose for the fixed COCO label set; n size runs in real time on CPU.
- In/out Image to Boxes, Keypoints, Scores
License
AGPL-3.0
Hardware
Runs on CPU
View modelJSON
Not runnable today
YOLO-World v2 (Ultralytics)
Detection · yolov8s-worldv2.pt
Open-vocabulary detection from a list of text class names, in real time; set classes per namespace without training.
- In/out Image, Text to Boxes, Labels, Scores
License
AGPL-3.0
Hardware
Runs on CPU
View modelJSON

Questions

Which of these models does Mixpeek run today?: 15 are served on Mixpeek as Ray Serve apps that start at zero replicas and wake on first use (Multilingual E5 Large Instruct, all-MiniLM-L6-v2, SigLIP Base patch16 224, InsightFace buffalo_l (SCRFD-10GF + ArcFace ResNet50 WebFace600K), InsightFace buffalo_sc (SCRFD-500MF + MobileFaceNet WebFace600K) and others). 13 are provider APIs Mixpeek calls by name. 34 can run as a custom plugin and do not run yet. 29 are blocked today; each page says why.
How is the default order decided?: By usage, the namespaces and collections that use each model, counted on 2026-11-06T13:42:25Z, then by how many templates use it, whether it runs today and whether we have our own evaluation.
What does unmeasured mean on a model?: We have not measured that model's cost or quality on our own image and data yet. Publisher benchmarks are linked from the model's sources and are not copied into the measured fields.
Can an agent read this catalog?: Yes. /models.json holds every record and /models/<id>.json holds one. They are the same records this page renders, generated from the catalog the Mixpeek server is built from.
How do I use a model in a namespace?: Open the model and deploy a template that already uses it, or add it to a new or existing namespace from the Designer in Studio. Every served model starts at zero replicas, so an idle namespace costs nothing for inference.

Hugging Face model directory

Directory pages for 422 public models. They are not in the catalog above, and Mixpeek does not run them as listed.