Visionular
Visionular (微帧科技) provides AI-powered video encoding and video AI services to global enterprise customers, offering self-developed AV1, H.265, and H.264 codec engines alongside AI enhancement, quality assessment, content analysis, and a new AI model aggregation platform delivered via cloud APIs and SDKs.
- Company typePrivate
- Founded2018
- HeadquartersLos Altos, United States
- Headcount51–100
- GTM typeB2B
- OfferingSoftware
What Visionular does
Visionular (微帧科技) is a global provider of AI-powered video encoding and video AI services founded in 2018 and headquartered in Hangzhou, China, with a sales office in Silicon Valley and an engineering office in the United Kingdom. The company builds self-developed video codec engines — Aurora-AV1 (the world's first commercial AV1 software encoder, covering VOD, live, and RTC), WZ265 (H.265/HEVC), and WZ264 (H.264/AVC) — each enhanced with AI preprocessing, content-aware (CAE) perceptual encoding, intelligent rate control, Per-Title encoding, and region-of-interest protection. Beyond core encoding, the product portfolio extends to an AI Ultra-High Definition Processing Engine (super-resolution, HDR, denoising, color enhancement), an Image Intelligent Compression Engine covering JPEG/WebP/HEIF/AVIF formats, a GPU hardware encoding optimization layer for NVIDIA NVENC achieving ~30% bitrate savings, a no-reference video quality assessment system (WZVQA), and a broad suite of AI services spanning content understanding (smart clipping, summarization, smart covers, target tracking), intelligent media (highlights, subtitle removal, landscape-to-portrait conversion, virtual advertising), and video security (Video DNA fingerprinting, blind watermark, content review).
The commercial model is hybrid: usage-based AuroraCloud API services for cloud transcoding, SDK licensing across mobile (iOS, Android, HarmonyOS NEXT), desktop, and server platforms on a subscription basis, and — most recently — an AI Large Model Aggregation Platform (launched May 2026) providing unified API access to multiple AI foundation models. Distribution is enterprise-direct through regional sales and solutions teams in North America (Silicon Valley) and Europe (UK), with global support across 10+ countries including China, the US, the UK, India, Singapore, the Philippines, Sweden, Brazil, Saudi Arabia, and Canada. The company reports serving 150-200+ enterprise customers with 90%+ product repurchase rate, processing 2 billion+ minutes of video per month, with named customers spanning broadcasting (CCTV, ABS-CBN), social and short video (Snap, Xiaohongshu, Pinterest, ShareChat, Josh, KUMU), live streaming and sports (Huya, FanCode, NBC Sports Engine Play, Hudl, Intinor), payments (Alipay), smart devices (Xiaomi TV, Tmall Genie), and infrastructure (H3C, Agora, Shopee, Baidu). Pricing is quote-based and not publicly disclosed, and total disclosed funding to date is approximately $8.9M across a 2018 seed round and a 2021 round.
Visionular firmographics
Firmographics- Name
- Visionular
- Legal name
- 微帧科技
- Website
- https://visionular.com
- Company type
- Private
- Founded year
- 2018
- Operating status
- Operating
- Headcount range
- 51–100 employees
- Short description
- Visionular (微帧科技) provides AI-powered video encoding and video AI services to global enterprise customers, offering self-developed AV1, H.265, and H.264 codec engines alongside AI enhancement, quality assessment, content analysis, and a new AI model aggregation platform delivered via cloud APIs and SDKs.
- Ownership category
- akta.pro rank
Visionular industry classification
Industry- Product category
- Video Encoding & AI Video Processing Software
- NAICS
- Other Motion Picture and Video Industries (512199), Motion Picture and Video Industries (5121)
- SIC
- Services-Motion Picture & Video Tape Production (7812)
- akta.pro primary industry
- Content-Aware/AI Media Processing (Per-Title Encode, Scene Detection) (MPAHAHAH)
- akta.pro secondary industries
- Video Creation, Editing & Publishing Tools (Creator Software) (MPACABAD), Creator Tools & Production (Overlays, Alerts, Encoding, Editing, Audio) (MPAFAEAC), AI Application Enablement Platforms (Copilot/Agent Frameworks, SDKs) (HDAEANAJ)
Keywords
Where Visionular is headquartered
LocationHeadquarters
- HQ city
- Los Altos
- HQ country
- United States
- HQ region
- North America
Offices3 records
Markets served
Visionular business model
Business model- GTM type
- B2B
- Offering type
- Software
- Cost components
- Technology or R&D, Personnel, Marketing or Sales, Infrastructure, Operations
Revenue model
- Cloud Video Encoding Services (AuroraCloud): Cloud-based video encoding and processing services provided via API. Customers upload video content and receive encoded output with billing based on video volume processed. No migration costs required for customers to integrate the service.
- SDK Licensing: Licensing of video encoding and enhancement SDKs (WZ264, WZ265, Aurora-AV1) for integration into customer applications and platforms. SDKs available for mobile (iOS, Android, HarmonyOS NEXT), desktop, and server environments.
- AI Model Aggregation Platform: Platform providing unified API access to multiple AI foundation models (LLMs, vision models) with automatic routing for optimal pricing. Revenue from API usage fees with direct connection to model providers.
Go-to-market motion3 records
Distribution channels3 records
Marketing channels7 records
Visionular product offering
Product offeringCore offering
Visionular builds and licenses self-developed AI-enhanced video encoding engines (Aurora-AV1, WZ265, WZ264) plus a suite of AI video services (ultra-HD processing, content understanding, video security) and the AuroraCloud API platform. These products are sold to enterprise video platforms, streaming services, media companies, and tech enterprises to reduce bandwidth and storage costs by 30–50% while enabling ultra-HD, low-latency, AI-powered video delivery.
Product overview
Visionular (微帧科技) is a global provider of AI-powered video encoding and video AI services. The product portfolio consists of three interconnected layers: (1) Core Video Intelligent Encoding Engines — spanning Aurora-AV1 (AV1), WZ265 (H.265/HEVC), and WZ264 (H.264/AVC) encoders plus a GPU hardware encoding optimization module, all enhanced with AI preprocessing, CAE perceptual encoding, intelligent rate control, and Per-Title optimization; (2) AI Image and Video Enhancement — an AI Ultra-High Definition Processing Engine (intelligent super-resolution, HDR, color enhancement, denoising), a Smart Terminal Imaging module for on-device real-time enhancement, an Image Intelligent Compression Engine supporting JPEG/WebP/HEIF/AVIF formats, and a WZVQA AI video quality assessment system; (3) AI Video AI Services — three suites for Intelligent Media (highlights, subtitle removal, L2P, virtual ads), Content Understanding (smart clipping, cover, summary, target tracking), and Video Security (Video DNA, blind watermark, content review); all delivered via AuroraCloud cloud services with API access, covering short drama, short video, live streaming, film/TV OTT, video conferencing, surveillance, cloud desktop, smart displays, and AIGC platforms. The company also offers an AI Large Model Aggregation Platform and the AuroraVision AI suite of AI video processing services.
Differentiator
Problem solved
Functional benefit
Products and services
- Video Intelligent Encoding Engine (Aurora-AV1 / WZ265 / WZ264 unified AI+ encoder platform) Unified AI-enhanced video encoding platform spanning Aurora-AV1 (AV1), WZ265 (H.265/HEVC), and WZ264 (H.264/AVC) self-developed encoders with AI preprocessing, CAE perceptual encoding, adaptive content recognition, ROI region-of-interest protection, Per-Title optimization, and intelligent rate control. Targets enterprise VOD, live, and RTC scenarios and achieves 30–50% bitrate savings vs open-source encoders at equivalent quality.
- AuroraCloud Cloud Video Processing Service Cloud-based API service for intelligent transcoding, multi-cloud deployment, adaptive scenarios, and dynamic resource scaling for high-concurrency transcoding. Supports short drama, short video, live streaming, film/TV OTT, video conferencing, surveillance, cloud desktop, and smart display workloads, delivering 30–50% bandwidth savings with no migration cost.
- AI Ultra-High Definition Processing Engine AI-driven suite for video quality enhancement covering intelligent super-resolution (480p to 4K+ upscaling), HDR and Pseudo HDR processing, color enhancement, intelligent deinterlacing, denoising, and sharpening. Applied to film restoration, live event enhancement, and premium content production for enterprise media and broadcast customers.
- Image Intelligent Compression Engine (WZ-JPEG / WZ-WebP / WZ-HEIF / WZ-AVIF) Deep learning-based image compression engine supporting WZ-JPEG, WZ-WebP, WZ-HEIF, and WZ-AVIF formats, using ROI detection and perceptual encoding to deliver 27–42% file size reduction versus open-source alternatives at similar VMAF quality. Supports 10-bit HDR encoding and 4x faster encoding speed than open-source encoders.
- GPU Video Hardware Encoding Optimization Engine AI-enhanced optimization layer for NVIDIA NVENC hardware encoding modules, integrating WZVQA quality assessment, CAE content-aware encoding, and quality enhancement to achieve approximately 30% bitrate savings versus standard NVENC without additional compute resources. Targets cloud GPU transcoding workloads at enterprise scale.
- WZVQA No-Reference Video Quality Assessment AI-powered no-reference video quality assessment system that outputs an overall quality score plus 10 sub-scores (blur, block noise, grain noise, overexposure, underexposure, contrast, aesthetic impact, color richness, sharpness, sandwich video detection) using probability representation, closely mimicking human visual perception.
- AI Video Security Suite (Video DNA / Blind Watermark / AI Content Review) AI-powered video security suite covering Video DNA fingerprinting for content identification and tracking, Blind Watermark for invisible copyright embedding and leak tracing, and AI Content Review for automated moderation of nudity, violence, and terror content with customizable regional policies.
- AI Intelligent Media Suite (Highlights / Subtitle Removal / L2P / Virtual Ads) AI media automation suite including AI virtual advertising insertion for sports/live broadcasts, landscape-to-portrait video conversion, AI subtitle removal with intelligent background inpainting, and automated highlight extraction. Used by enterprise media, sports, and entertainment platforms for content production.
- AI Content Understanding Suite (Smart Clipping / Smart Cover / Smart Summary / AI Target Tracking) Multi-modal AI content analysis suite covering smart video clipping, smart cover generation, AI video summarization, and CNN-based target tracking to power content production and platform operations for video enterprises.
- Smart Terminal Imaging On-device real-time image enhancement technology running on smart device GPUs and NPUs, enabling super-resolution, image enhancement, and color correction at the terminal. Deployed in Tmall Genie smart displays, Xiaomi TV, and automotive infotainment systems.
- AI Large Model Aggregation Platform AI foundation model aggregation platform that provides a single API to access multiple LLMs and vision models with auto-routing for optimal pricing. Aims to reduce AI application development costs by 30% via direct model sourcing and supports global compliance and deployment for enterprise and AIGC platform customers.
- AuroraVision AI Product Series Comprehensive AI video processing services suite integrating AI VQA quality assessment, intelligent frame interpolation, super-resolution and enhancement, and AI ball/object tracking. Covers the full video processing pipeline from content production through quality enhancement to security distribution.
Quantifiable outcome
- 30%-50% bitrate savings vs open-source encoders
- +6 more outcomes
Companies that use Visionular
Customer profileNamed customers18 records
Segments10 records
Ideal customer profiles2 records
Visionular technology and API
TechnologyTechnology focussed Yes
API detail
- Has API
- Yes
- API docs
- API detail
Core technology
AI maturity
App detail
Integration3 records
AI capability9 records
Feature15 records
Visionular partnerships and signals
Strategic signalPartnerships
Eleven partnerships are on record, tiered core and flagship.
- PAGC 2026 Conference OrganizercoreExhibited at PAGC 2026 showcasing short drama AI solutions and AI model aggregation platform. Won 2026 H1 AI Globalization Pioneer Award.
- ABS-CBN (菲律宾最大广播电视媒体集团)coreProviding AuroraCloud cloud video compression services for iWant global streaming platform. Achieved 40%+ bandwidth reduction. Partnership enables ABS-CBN to provide premium video experiences to global audiences.
- IntinorcoreIntegrated WZ265 video encoding into Intinor's Direkt link and Direkt router platforms for broadcast-quality streaming. Achieved professional-grade video transmission at only 2 Mbps for 4G/5G networks. Partnership enabling HD live streaming in challenging network environments.
- NBC Sports Engine PlayflagshipFull-stack support for NBC Sports Engine Play platform including competition live streaming, replay viewing, highlight reel production, real-time scoreboard updates, and unlimited cloud storage. Integration of efficient video AI services for stable HD sports streaming.
- UWA联盟coreParticipated in authoring 'Mobile Scene Extended Reality Ultra-High Definition Technology White Paper' led by UWA, CCTV, and Huawei. Contributed expertise in spatial video coding optimization.
- Agora (声网)coreRTC screen content encoding solution deployed at Agora and multiple RTC vendors. Aurora-AV1 shows absolute advantages in screen content encoding scenarios with 81.25% BD-rate gain vs x264.
- Tmall Genie (天猫精灵)coreJoint development of custom AI intelligent enhancement technology 'Jingling Original Picture' deployed on Tmall Genie smart display terminals. Leverages GPU and NPU for real-time visual enhancement. Breakthrough in terminal-side real-time video enhancement.
- NVIDIAcoreStrategic collaboration with NVIDIA for NVENC hardware encoding optimization. Integrating VQA assessment, CAE content-aware encoding, and quality enhancement with NVIDIA NVENC module. Achieves 30% bitrate savings vs standard NVENC while maintaining quality.
- HudlcoreIntegrated WZ264 and WZ265 encoding engines into Hudl sports analysis platform. Significant video quality improvement at reduced bitrates for professional and amateur sports team video analysis.
- AMDcoreCollaboration at MWC LA 2021 to demonstrate Aurora encoder achieving 8K AV1 Live on AMD EPYC 7763 (Milan) server.
- AOMedia (Alliance for Open Media, 开放媒体联盟)coreMember of AOMedia consortium since its founding. Zoe Liu contributed to AV1 standard development as Google AV1 team core algorithm engineer. Company recognized by AOM as first to achieve comprehensive AV1 software encoding support across all use cases (VOD, live, RTC). Active participation in AV2 standard development.
Scale indicators10 records
Recent moves6 records
Expansion highlights7 records
Visionular competitors and assessment
Company assessmentDirect peers
- Encoding.com (now part of JW Player): Encoding.com is a long-standing cloud video transcoding service offering API-based encoding at scale. Directly comparable as a cloud video processing API targeting enterprise media customers with similar encoding-as-a-service positioning.
- Mux: Mux offers a video API platform with encoding, streaming, and analytics for developers and enterprises. Competes directly with Visionular's AuroraCloud in cloud video transcoding and developer-first API delivery model.
- MainConcept: MainConcept provides professional video and audio codec SDKs (H.264, H.265, AV1, etc.) for integration into broadcast, streaming, and consumer applications. Highly comparable to Visionular's WZ264/WZ265/Aurora-AV1 SDK licensing business.
- Bitmovin: Bitmovin provides cloud-based video encoding and player SDKs serving streaming platforms globally. Direct competitor in API-first video infrastructure, with overlapping enterprise customer segments and similar per-title/per-shot encoding optimization capabilities.
- Beamr Imaging: Beamr offers GPU-accelerated video optimization technology (Beamr 5x) delivering bitrate savings at H.264/HEVC quality levels. Directly comparable AI-enhanced video encoding approach, also positioned around bandwidth savings for streaming platforms.
Broad incumbents
- MediaKind: MediaKind (formerly Ericsson Media Solutions) offers media processing, encoding, and delivery platforms for broadcasters and operators. Comparable in serving major broadcast customers (NBC Sports, ABS-CBN segments) with end-to-end video processing infrastructure.
- AWS Elemental MediaConvert: AWS Elemental MediaConvert is a cloud-based file-based video transcoding service from Amazon, offering H.264, HEVC, and AV1 encoding at scale. Represents the hyperscaler encoding competitive threat that Visionular must differentiate against on compression efficiency and AI-enhanced features.
- Ateme: Ateme provides video compression and delivery solutions for broadcast, OTT, and streaming applications, with notable expertise in HEVC and AV1. Comparable in serving sports streaming and broadcast customers (FanCode, ABS-CBN overlap) but with broader portfolio and larger installed base.
- Harmonic: Harmonic is a major broadcast video infrastructure vendor providing encoding, processing, and delivery solutions for OTT and pay-TV operators. Overlaps with Visionular's OTT/streaming and broadcast customer use cases, but operates at much larger scale with hardware-software bundled offerings.
Regional players
- Tencent Cloud Media Processing Service: Tencent Cloud MPS provides cloud-based media transcoding, AI enhancement, and content delivery services primarily across Asia. Comparable in serving Chinese tech and media companies with similar AI video processing capabilities, representing regional cloud platform competition.
Market position
Strengths4 records
Weaknesses4 records
Competitive moat5 records
Key risks6 records
Key highlights7 records
Customer concentration
Visionular social profiles
Digital presenceVisionular financial estimates
Financial estimateRevenue estimate
Valuation estimate
Visionular leadership team
Management profileNumber of profiles
Profiles6 records
Visionular funding detail
Funding detailFunding overview
Funding rounds3 records
Investors
Funding detail is available on the Subscription and Enterprise plan.Contact sales →
Visionular M&A and investment
M&A and investmentM&A
Investments
M&A and investment is available on the Subscription and Enterprise plan.Contact sales →
Frequently asked questions about Visionular
What does Visionular do?
Visionular builds and licenses self-developed AI-enhanced video encoding engines (Aurora-AV1, WZ265, WZ264) plus a suite of AI video services (ultra-HD processing, content understanding, video security) and the AuroraCloud API platform. These products are sold to enterprise video platforms, streaming services, media companies, and tech enterprises to reduce bandwidth and storage costs by 30–50% while enabling ultra-HD, low-latency, AI-powered video delivery.
Is Visionular a public or private company?
Visionular is a private company. It is classified as venture growth investor backed and is currently operating.
When was Visionular founded?
Visionular was founded in 2018. It employs 51 to 100 people.
Where is Visionular based?
Visionular is headquartered in Los Altos, United States, in the North America region.
How does Visionular make money?
Three revenue lines are on record. Cloud Video Encoding Services (AuroraCloud) is the primary driver. The others are SDK Licensing and AI Model Aggregation Platform.
Who are Visionular's main competitors?
Direct peers on record are Encoding.com (now part of JW Player), Mux, MainConcept, Bitmovin and Beamr Imaging. Broad incumbents are MediaKind, AWS Elemental MediaConvert, Ateme and Harmonic. Tencent Cloud Media Processing Service is listed as a regional player.
Does Visionular have an API?
Yes. Visionular offers cloud-based API services for video encoding, AI processing, and cloud media processing. Clients can access encoding capabilities without custom API development through S3 event triggers that automatically initiate transcoding upon upload, supporting dynamic resource scaling for high-concurrency needs. Access to cloud infrastructure is available in AWS Southeast Asia region nodes, with multi-cloud deployment capabilities.
What industry is Visionular in?
Visionular's product category is Video Encoding & AI Video Processing Software. Its primary akta.pro industry code is MPAHAHAH, Content-Aware/AI Media Processing (Per-Title Encode, Scene Detection), with a secondary code of MPACABAD, Video Creation, Editing & Publishing Tools (Creator Software). Its NAICS code is 512199 and its SIC code is 7812.