Sighthound AI
sighthound.com
- Free plan
- Windows
- macOS
- Linux
- Web
- iPhone and iPad
- Android
- API
- Self-hosted
What Sighthound AI is
Sighthound AI makes computer vision products for vehicle recognition, media redaction, and edge AI hardware. Retriever reads license plates and identifies a vehicle’s make, model, color, and generation from live camera streams or uploaded footage. It runs on customer machines through Docker Compose, works in a browser, and can operate offline without a required cloud connection. It does not provide vehicle owner or location information. Retriever supports almost any IP camera, though Sighthound recommends specialized LPR cameras for fast-moving vehicles. Redactor detects and redacts video, image, and audio content, and creates searchable transcripts to find keywords for redaction in 8+ languages. Its integration options include a REST API, batch jobs, webhooks, and a brandable embedded editor. Sighthound says media files stay local and supports local, on-premise, and private cloud deployments. Redactor is described as built for GDPR, CCPA, FOIA, HIPAA, and other global privacy regulations. Retriever’s listed uses span parking, enforcement, law enforcement, tolls and traffic, fleets, logistics, retail, and automotive. Pricing is on request; a free plan and trial are available.
Who it suits
Sighthound AI suits teams handling vehicle recognition or sensitive media, including the listed parking, enforcement, fleet, logistics, retail, and automotive use cases. Retriever is also relevant to organizations that need local or offline operation and support for IP cameras.
Good to know, and what to watch
- Retriever can work offline without a cloud connection.
- Redactor handles video, image, and audio content.
- Searchable transcripts support keyword redaction in 8+ languages.
- Redactor offers API, batch, webhook, and editor integrations.
- A free plan and trial are available.
- Retriever does not provide vehicle owner or location information.
- Fast-moving vehicles may call for specialized LPR cameras.
- Sighthound Video no longer receives active feature development or regular updates.
- Pro costs $2,500 a year.
Plans
Every plan and priceEnterprise
Free
Pro
Server
Pricing on request
Getting started
- Visit https://www.sighthound.com/ to explore Sighthound AI.
- Choose the Free, Pro, or Server plan; Enterprise uses custom pricing.
- For Retriever, install on customer machines through Docker Compose or run it in a browser.
- Use a compatible IP camera or upload footage for Retriever; specialized LPR cameras are recommended for fast-moving vehicles.
- For Retriever installation help, use the Developer Portal or Knowledge Base, or contact Sighthound directly.
Limits to check
The Free plan is limited to 1 user, 24-hour access, and up to 2 minutes, and is hosted by Sighthound. Pro costs $2,500 a year and includes 1 user; Server costs $3,500 a year, starts with 1 user, and permits additional users up to 5 at $500 a year per user.
The Club's review
Sighthound AI brings together vehicle recognition and media redaction for organizations handling camera feeds or sensitive recordings. It is strongest for teams that need local processing, structured redaction workflows, or vehicle identification at the edge. Its two main products do different jobs, so the right fit depends on whether the priority is identifying vehicles or preparing media for controlled sharing.
Overview
Retriever reads license plates and identifies a vehicle’s make, model, color, and generation from live camera streams or uploaded footage. Redactor detects and redacts video, image, and audio content, and can create searchable transcripts to find keywords in 8+ languages. Together, they address operational computer-vision work rather than general-purpose image recognition.
Retriever identifies vehicles but does not provide owner or location information. That makes it useful for parking, tolls, traffic, fleet, logistics, retail, automotive, enforcement, and law enforcement workflows that need vehicle attributes, while leaving ownership and whereabouts to other systems.
Key features
Retriever for camera workflows
Retriever installs on customer machines through Docker Compose, runs in a browser, and works offline without a required cloud connection. This gives teams control over where video is processed and keeps operation from depending on a cloud connection. It supports almost any IP camera; for fast-moving vehicles, Sighthound recommends specialized LPR cameras, an important consideration where capture speed affects recognition.
Installation help is available through the Developer Portal and Knowledge Base, with direct contact as another support route. That gives deployment teams a path to assistance, though the product’s stated support channels are not a substitute for checking operational requirements before rollout.
Redactor for sensitive media
Redactor handles video, images, and audio, with searchable transcripts that help locate keywords for redaction in 8+ languages. REST API access, batch jobs, webhooks, and a brandable embedded editor make it suited to repeatable processing pipelines as well as workflows where reviewers need an editing interface.
Sighthound says media files stay within the chosen environment, with local, on-premise, and private cloud deployments supported. Its stated compliance focus includes GDPR, CCPA, FOIA, HIPAA, and other global privacy regulations. Those deployment and workflow options are meaningful for organizations with strict media-handling needs, but the claims do not establish that any particular deployment meets an organization’s own compliance obligations.
Object detection, face recognition, custom models, and hybrid deployment are also included in Sighthound AI’s product scope. Sighthound Video is a separate legacy product: it is open source, no longer receives active feature development or regular updates, and receives best-effort support for critical issues.
Pricing
The Free plan costs nothing and allows one user, 24-hour access, up to 2 minutes, and Sighthound-hosted use. It offers a way to try a short workflow, but the two-minute cap makes it unsuitable for sustained video processing.
Pro costs $2,500 a year, billed $2,500 /year. It supports one user, unlimited video length, desktop installation, and email support. This is the clearest fit for a single user who needs longer video processing and a desktop install; the annual price and one-user limit make it a substantial commitment for an individual workflow.
Server costs $3,500 a year, billed $3,500 /year. It starts with one user, with additional users up to five available at $500/yr/user, and includes server deployment, email support, and product training. Teams that need server deployment and training get a more suitable package than Pro, while the five-user ceiling may constrain larger groups.
Enterprise has custom pricing for one or more users and includes API access, a dedicated private Slack, and private deployment or OEM options. It is aimed at organizations that need those integration or deployment terms; teams should weigh the custom quote against the fixed annual plans. A free plan and free trial are available, though only the Free plan’s stated usage cap is clear.
Platforms
Sighthound AI lists Android, iOS, Linux, macOS, Windows, web, API, and self-hosted platforms. Retriever’s Docker Compose installation and browser interface, together with its offline operation, make local deployment a particular strength for camera environments.
Who it's for
Choose Sighthound when vehicle attributes or privacy-conscious media redaction are core needs, especially if processing must stay local or on-premise. Retriever fits parking, traffic, toll, fleet, logistics, retail, automotive, and enforcement operations that need vehicle recognition from camera footage. Redactor better suits organizations managing sensitive recordings and building repeatable review pipelines. Look elsewhere if the requirement is vehicle owner or location information, or if the task is broad image recognition without these specialized workflows.
Pros and cons
- Pro: Retriever works offline on customer machines and supports a wide range of IP cameras, giving deployments flexibility where cloud access is unsuitable.
- Pro: Redactor combines audio, image, and video redaction with searchable multilingual transcripts and integration tools for automated workflows.
- Pro: Local, on-premise, and private cloud options address organizations that need tighter control over sensitive media.
- Con: Retriever does not return owner or location information, limiting it to vehicle recognition rather than broader investigative or tracking needs.
- Con: The Free plan stops at two minutes and 24-hour access, while Pro begins at $2,500 a year for one user.
- Con: Fast-moving vehicles may call for specialized LPR cameras, so an existing camera setup may not be the best fit for every capture environment.
- Con: Sighthound Video no longer receives regular updates or active feature work, making it a poor choice for users seeking a maintained legacy video product.
Alternatives
AI Image Recognition Software is a broader category to browse when the task does not center on Sighthound’s vehicle-recognition or media-redaction workflows.
Face++ is worth considering for teams seeking a freemium API with a free key and multiple supported platforms.
Google Cloud Vision API may suit image-processing workflows that prefer a pay-as-you-go API model and monthly usage tiers.
SentiSight.ai is an alternative for teams seeking a freemium computer-vision service with an offline model license option.
Yandex Vision OCR is an option for readers comparing paid OCR and license plate recognition plans.
CompreFace may fit teams looking for a free, open-source, Docker-based deployment with CPU and GPU support.
InsightFace is a free option for research and non-commercial projects, with commercial model use requiring a separate license.
MediaPipe is a free alternative for teams looking for on-device machine-learning solutions and a framework.
Roboflow may suit model-training and inference workflows, with a free tier of 10 monthly credits and a Core plan at $39 a month.
Verdict
Sighthound AI is a strong choice for organizations that need local vehicle recognition or privacy-oriented media redaction with practical integration options. Its clearest advantages are Retriever’s offline camera workflows and Redactor’s controls for processing sensitive media. Choose another tool if you need vehicle ownership or location data, or if Sighthound’s annual Pro and Server costs do not fit the scale of your work.
The verdict
Sighthound AI combines vehicle recognition with media redaction, including local deployment options and integration tools. Check Retriever’s information limits and the listed plan terms against your needs; Enterprise pricing is on request.
Questions people ask
Is there a free plan?
Yes. The Free plan costs nothing and includes 1 user, 24-hour access, up to 2 minutes, and hosting by Sighthound.
What do the paid plans cost?
Pro is $2,500 a year and Server is $3,500 a year. Enterprise has custom pricing.
What platforms does Sighthound AI support?
Listed platforms are Android, API, iOS, Linux, macOS, self-hosted, web, and Windows.
Can Retriever work without a cloud connection?
Yes. It runs offline without a required cloud connection and can be installed on customer machines through Docker Compose or used in a browser.
Does Retriever identify vehicle owners or locations?
No. Sighthound says Retriever does not provide vehicle owner information or location information.
Is Sighthound Video actively developed?
No. Sighthound Video is open source but no longer receives active feature development or regular updates; support is best effort for critical issues.
What the maker publishes
- Audio redaction
- Redactor can mute, beep or scramble speech in audio and video files. source
- Audio transcription
- Redactor creates searchable transcripts for keyword redaction in more than eight languages. source
- Company
- The maker identifies itself as Sighthound, Inc. and lists co-founder Jonathan Taylor as Executive Chairman, co-founder Dr. Zain Masood as CTO, and co-founder Markus Hahn as Chief Architect. source
- Company founded
- LinkedIn states that Sighthound was founded in 2013. source
- Company headquarters
- Crunchbase lists headquarters in Winter Park, Florida, United States. source
- Company legal name
- Crunchbase identifies the legal entity as Sighthound, Inc. source
- Company size
- Crunchbase lists Sighthound as having 11–50 employees. source
- Compliance
- Sighthound says Redactor is built for GDPR, CCPA, FOIA, HIPAA, and other global privacy regulations. source
- Compliance positioning
- Sighthound says Redactor can support workflows involving HIPAA, GDPR, CCPA/CPRA, FERPA, FOIA and CJIS, while stating that compliance remains the customer's responsibility. source
- Compute Camera
- The current Compute Camera version 4 uses an 8.3MP Sony STARVIS sensor, offers motorized lens options, up to 4TB NVMe storage and IP67 protection. source
- Compute Node
- The current Compute Node version 5 supports existing RTSP cameras, up to four streams in Super Mode and NVIDIA Orin NX or Nano hardware. source
- Customer fit
- Sighthound describes Retriever use cases including parking, enforcement, law enforcement, tolls and traffic, fleets, logistics, retail, and automotive. source
- Deployment
- Redactor supports desktop, server, private-cloud, offline and air-gapped deployments, and customer-controlled AWS, Azure or Google Cloud environments. source
- Developer toolkit
- The Redactor Agent Toolkit contains Docker Compose configurations, TypeScript definitions, AI-agent documentation and an OpenAPI specification. source
- Device compatibility
- Retriever can work with almost any IP camera, while Sighthound recommends specialized LPR cameras for fast-moving vehicles. source
- Embedded integration
- Redactor provides an embeddable, brandable editing interface for integrating redaction into another product. source
- Funding record
- Crunchbase lists seed, Series A and convertible-note funding rounds and names Entrenext Ventures among the investors. source
- Hardware
- Sighthound Compute devices run AI locally with NVIDIA GPUs, Linux, Docker support, RTSP camera compatibility and APIs for third-party integration. source
- Human review limit
- Sighthound says Auto Detect is a first pass and recommends human review and correction before exported footage is released. source
- Image redaction
- Redactor can blur, pixelate or fill sensitive content in photographs and screenshots. source
- Legacy open source
- Legacy Sighthound Video is now open source, is no longer under active feature development and receives only best-effort critical-issue support. source
- Legacy product status
- Sighthound Video is open source and no longer receives active feature development or regular updates; support is best effort for critical issues. source
- Legacy system requirements
- Sighthound Video 7.0.16 requires macOS 10.15 Catalina or later or Windows 10 or later; the page notes a macOS audio-playback crash issue. source
- License-plate recognition
- Retriever reads plates from many countries and reports plate characters and region information for the United States, Canada and major European countries. source
Checked against the maker's pages on 21 Oct 2026.