London-built AI avatar video platform used by 70% of FTSE 100 companies
Synthesia is a London-based AI video generation platform enabling users to create professional videos with AI avatars and synthetic voices across 140+ languages without cameras, studios, or actors. Founded in 2017 by Victor Riparbelli, Steffen Tjerrild, Matthias Niessner, and Lourdes Agapito, the company reached a $4 billion valuation in October 2025 with $100M+ ARR. Used by 70% of FTSE 100 and 90%+ of Fortune 100 companies.
Headquarters
London, United Kingdom
Founded
2017
Pricing
EU Data Hosting
No
Employees
201-500
Free
$29/mo
$89/mo
Contact Sales
Billing: monthly, annual
By October 2025, Synthesia had reached a $4 billion valuation with more than $100M in annual recurring revenue. Those numbers matter for context because Synthesia is not a well-funded startup experimenting with AI video. It is the market-defining platform in enterprise synthetic media, used by 70% of FTSE 100 companies and cited by Gartner and Forrester as the category leader.
The company was founded in 2017 by Victor Riparbelli (CEO, Danish), Steffen Tjerrild, Matthias Niessner (an AI professor at TU Munich), and Lourdes Agapito (a computer vision professor at UCL). This academic origin is visible in the product: Synthesia has consistently led on technical capability rather than growth-hacking its way to adoption. The Series E in 2025 (led by GV with Adobe Ventures participating) brought total funding to over $200M.
The registered entity is Synthesia Limited, incorporated at Companies House (number 10571132) with a London registered office. The UK classification as european rather than EU-member is consequential for compliance purposes. UK-GDPR mirrors EU GDPR substantively but sits outside the EU data protection framework, which matters for regulated organisations with EU data residency requirements.
The core product generates professional video from a text script, selecting from 230+ AI avatars speaking in 140+ languages. No camera, studio, lighting rig, or human performer is required. The use case that drove initial enterprise adoption was corporate training: L&D teams can produce or update training videos by editing a script, eliminating re-recording cycles.
Synthesia's avatar library stands at 230+ stock avatars, with representation across demographic groups, presentation styles, and professional contexts. The Express-2 generation model, released in 2025, significantly improved the naturalness of avatar movement and lip-sync accuracy compared to earlier generations, reducing the uncanny valley effect that made earlier AI video immediately identifiable.
Custom Personal Avatars are created from a 30-minute consent-verified recording session. The result is a synthetic presenter built from the user's own likeness, speaking any script in any supported language with the user's voice characteristics. This costs approximately £800 as a one-time fee (included in the Starter plan and above). The consent verification process is mandatory and explicit: Synthesia does not generate avatars without documented consent from the depicted person.
The 140+ language coverage is comprehensive across European, Asian, Middle Eastern, and Latin American markets. Voices are synthetic but styled for naturalness in the target language; Synthesia has invested significantly in prosody and intonation beyond direct translation-then-speech.
AI dubbing extends this capability to existing video: upload an English video, and Synthesia can generate a dubbed version in another language with the avatar's mouth movements resynchronised to the new audio track. For companies running global learning programmes, this reduces localisation cost and cycle time considerably. Contrast this with tools like DeepL, which handles text translation at very high quality; Synthesia operates at the video layer, where the translation and the visual delivery are inseparable.
The Interactive AI video agent feature (available on Enterprise plans) is Synthesia's most significant capability evolution beyond static content generation. Rather than producing a single linear video, it creates branching video experiences where viewer responses (entered via text or button selections) determine which video path plays next, with an AI avatar conducting the conversation.
The practical applications span sales enablement (personalised product demos), learning programmes (adaptive quizzes with video feedback), and customer service (video-based self-service flows). This moves Synthesia from a content creation tool toward an AI communication infrastructure product, which explains the $4B valuation relative to revenue.
In September 2024, Synthesia became the first generative AI company globally to achieve ISO/IEC 42001 certification, the international standard for AI management systems. This certification covers governance of AI development, deployment, and monitoring, going substantially beyond the data security assurance of ISO 27001.
Synthesia also embeds C2PA (Coalition for Content Provenance and Authenticity) content credentials in every generated video. These machine-readable metadata tags allow any C2PA-aware platform or tool to verify that content was AI-generated by Synthesia, providing a technically robust mechanism for provenance verification. This is a direct response to deepfake concerns and positions Synthesia constructively within the EU AI Act's requirements around transparency for synthetic media. For European enterprises navigating AI Act compliance, this credential infrastructure is a meaningful differentiator.
Synthesia's API allows programmatic video generation at scale, useful for personalised outreach campaigns (generating hundreds of unique videos with name and detail variations) or for integrating video generation into product workflows. The API covers full video creation including avatar selection, script, language, and template.
The free plan covers 3 minutes per month across 9 avatars, with Synthesia watermarks on output. It is sufficient for testing avatar quality and script rendering but not for production evaluation.
Starter at $29/month ($18/month on annual billing) gives 10 minutes per month, 125+ avatars, one Personal Avatar creation, and removes the watermark. The 10-minute monthly allowance is tight for teams producing regular content, but sufficient for organisations creating one or two training or marketing videos per month.
Creator at $89/month ($69/month annually) increases the allowance to 30 minutes, expands the avatar library to 180+, and adds brand kit functionality. This is the operative tier for most marketing and L&D teams with regular video production needs.
Enterprise pricing is custom. It includes unlimited video generation, interactive AI video agents, SSO/SAML, dedicated customer success, and formal compliance documentation (SOC 2, ISO 27001 reports). For organisations with L&D programmes at scale, the per-seat economics at Enterprise tier are typically more favourable than the per-minute model at lower tiers.
There is no free trial for paid tiers; evaluation happens via the free plan.
Synthesia Limited is a UK-incorporated company, placing it under UK-GDPR rather than EU GDPR. UK-GDPR mirrors EU GDPR in substance following the UK's departure from the EU, and the UK has been granted an EU adequacy decision, meaning personal data can flow from the EU to UK entities without additional safeguards in most circumstances.
The practical implication for European enterprise buyers: Synthesia is not an EU-member company, and data processing may not occur within EU borders by default. Organisations with strict EU data residency requirements should verify data hosting terms with Synthesia's enterprise team before signing.
On the certification side, Synthesia's posture is strong. SOC 2 Type II, ISO 27001, and the globally pioneering ISO/IEC 42001 AI management certification provide a robust audit trail. The C2PA content credentials and mandatory consent process for custom avatars address the synthetic media governance requirements that the EU AI Act introduces for high-risk AI systems.
For European AI infrastructure alternatives, Mistral AI provides EU-hosted language model inference that can be combined with Synthesia's avatar layer, though these are complementary rather than competing products. Buyers comparing Synthesia against the broader AI video generation category will find it the only EU-built option at this scale; the other tools on most shortlists, including HeyGen, are US-headquartered.
If your L&D team produces corporate training content at any volume, Synthesia's SCORM export, branching video agent capability, and brand kit make it the most capable platform in the category. The absence of cameras and studios removes a significant production bottleneck.
If you need multilingual video at scale, the AI dubbing feature reduces localisation cost compared to hiring voice actors and video editors per language. Translating a core English curriculum into 12 languages, for instance, can be done without re-recording.
If strict EU data residency is a hard requirement (not just a preference), the UK incorporation means you need to verify Synthesia's data hosting terms carefully, and may find an EU-member provider more appropriate.
If budget is the primary constraint, the Starter plan's 10-minute monthly allowance may prove too limiting for meaningful production use; the Creator tier at $89/month is where most functional teams actually operate.
Synthesia is the best-established platform in enterprise AI video generation, with technical credibility (academic founders, ISO 42001 certification), enterprise adoption (FTSE 100/Fortune 100 penetration), and a product that has evolved from static avatar generation toward interactive AI video agents. For European organisations comfortable with UK-GDPR governance, it is the default choice in the category.
The compliance caveat is genuine: UK incorporation is not the same as EU-member status, and the data residency question needs explicit verification for regulated sectors. Within those parameters, Synthesia's combination of avatar quality, language coverage, interactive capabilities, and content governance infrastructure is without peer in European-origin AI video tools.
Synthesia Limited is incorporated in the United Kingdom and operates under UK-GDPR, which mirrors EU GDPR substantively. The UK holds an EU adequacy decision, allowing personal data transfers from the EU. Enterprise customers should verify data residency in their contract, as default hosting may not be within EU borders. Synthesia holds SOC 2 Type II and ISO 27001 certifications.
Synthesia was the first generative AI company globally to achieve ISO/IEC 42001 certification for AI management systems. Its C2PA content credentials on generated videos provide machine-readable provenance. The platform uniquely offers interactive AI video agents that create branching, personalised video conversations rather than static content.
Synthesia supports 140+ languages and voice styles, including regional accents and natural prosody. All paid plans include full language access. The platform also offers AI dubbing, which converts existing English video content into other languages with synchronised avatar lip movements.
Custom Personal Avatar creation costs approximately £800 as a one-time fee and requires a 30-minute consent-verified recording session. The Starter plan includes one Personal Avatar creation within the subscription price. Enterprise plans include custom avatar creation within contract terms.
Yes. Synthesia exports SCORM-compatible packages for LMS integration and supports branded templates, video analytics, and script-based updates (edit the text to regenerate a video without re-recording). The Enterprise plan's interactive AI video agents enable adaptive learning paths with video-based branching. FTSE 100 L&D teams are a primary user profile.
London-built browser-based AI video editor with avatars, subtitles, and translation