Synthesia
AI avatar video for training and corporate comms, with brand controls and 140+ languages
Synthesia turns scripts, slides, documents or a chat prompt into narrated videos with AI avatars, in 140+ languages. Its 2026 releases add the Express-3 avatar engine, a conversational Assistant, Dubbing 2.0, AI B-roll, an MCP server for ChatGPT and Claude, and interactive Sessions for roleplay and surveys. It remains the enterprise-first option: brand kits, approvals, SSO and SCORM export.
- 1 creator using
- 2 news items tracked
- 2017 first launched
- Freemium pricing
- Easy learning curve
What Synthesia can do
- Infinite canvas No
- AI agent Partial
- Node / flow builder No
- Public API Yes
- MCP server Yes
15 of 26 capabilities confirmed
| Capability | Supported | Details |
|---|---|---|
| Text to image | Not confirmed | |
| Image editing | No | |
| Text to video | Partial | Avatar video from script is core; AI B-roll (Flux 3 up to 20s) is the generative-video part |
| Image to video | Not confirmed | |
| Video to video | Partial | AI Dubbing and translation re-voices and lip-syncs existing video |
| Native audio in video | Yes | Voice and lip sync are generated in the same pass; Flux 3 B-roll ships with synced audio |
| Lip-sync | Yes | Express-3 avatars (July 2026) with improved mouth tracking |
| Voice / TTS | Yes | AI voiceovers in 140+ languages and dubbing |
| Music & SFX | Not confirmed | |
| Upscaling | Not confirmed | |
| Character consistency | Yes | Personal (custom) avatars: 3 on Starter, 5 on Pro, unlimited on Enterprise |
| Camera control | No | |
| Custom training | Partial | Custom personal avatars from a photo or Avatar Builder; no LoRA or model training |
| Third-party models | Partial | B-roll uses third-party models: Flux 3 (Black Forest Labs) |
| Capability | Supported | Details |
|---|---|---|
| Infinite canvas | No | |
| Node / flow builder | No | |
| AI agent | Partial | Assistant (July 2026) drafts a full video from a prompt or document; it is a creation assistant rather than a multi-step planner that storyboards and generates unattended |
| Chat assistant | Yes | Conversational Assistant inside the editor; MCP connector also lets ChatGPT and Claude create videos |
| Team collaboration | Yes | Brand kits, approval controls, shareable links, team avatar sharing on Enterprise |
| Capability | Supported | Details |
|---|---|---|
| Public API | Yes | Video creation, dubbing, Interactive Avatar API and Billing API |
| MCP server | Yes | Synthesia MCP for ChatGPT and Claude (public beta) |
| Mobile app | Not confirmed | |
| Desktop / local | No | |
| Open source | No |
| Capability | Supported | Details |
|---|---|---|
| Free tier | Yes | Basic plan at $0: 500 credits, 10 min of video or dubbing per month, watermark applied |
| Commercial use | Yes | Stock avatars are commercially cleared; paid plans for commercial output |
Yes Partial No Not confirmed by the vendor
What makes Synthesia different
- 01
Commercially licensed stock avatars and a brand kit with approval controls, which is the main reason legal and HR teams pick it over open-ended video models
- 02
Paid plans remove the logo watermark from Starter upward (Basic free tier carries it), so the output is client-ready without a fee
- 03
Built-in SCORM export and LMS-friendly delivery, which HeyGen and most generators do not prioritise
- 04
Speaker-to-camera avatars with 140+ languages and AI dubbing that re-voices existing video, useful for localising a spokesperson once across markets
- 05
Sessions (survey and roleplay avatars) and an Interactive Avatar API cover conversational use cases that pure video generators do not
Latest from Synthesia
- May 15 AI in Marketing: The 20 Best AI Marketing Tools & Strategies (2025) Synthesia has released a curated guide to 20 AI marketing tools and workflows designed to automate creative asset management and campaign production for 2025.
- May 15 How to Create Training Videos: A Step-by-Step Guide Synthesia released a structured workflow for producing training videos, focusing on audience mapping and objective-driven AI video generation.
About Synthesia
Synthesia is still built for internal and corporate video: training modules, compliance explainers, product onboarding and multi-language localisation. Its avatar library is commercially licensed, and the paid tiers remove the logo watermark, so the output can go straight into a client or LMS deck.
What changed in 2026: the Express-3 avatar engine went live on every plan in July, with better lip sync, sentiment-driven expressions and roughly twice the render speed. The same launch added a conversational Assistant that turns a prompt or document into a draft video, and Dubbing 2.0 that holds sync across fast cuts and multi-person scenes. In September Synthesia added AI B-roll (Flux 3 from Black Forest Labs, up to 20 seconds with synced audio), an MCP connector for ChatGPT and Claude, and a Billing API for credit tracking.
For a filmmaker, Synthesia is a talking-head and explainer tool, not a cinematic generator. Use it when the shot is a presenter to camera, a localised version of an existing spokesperson video, or a fast explainer where legal review and brand consistency matter more than creative range. For storytelling footage, a text-to-video model like Veo or Kling is the better fit.
What's new in Synthesia
-
Synthesia MCP connector for ChatGPT and Claude
Generate a Synthesia video from inside ChatGPT or Claude, then open the draft in Synthesia to edit. Public beta on all paid plans.
Source: synthesia.io ↗ -
AI B-roll from Flux 3, with 720p/1080p/4K options
Generates up to 20 seconds of B-roll with synced audio, using Black Forest Labs Flux 3, and adds resolution choices for generated B-roll.
Source: synthesia.io ↗ -
Billing API and credit top-ups
A Billing API endpoint reports credit usage by breakdown, and users can buy credits or set automatic top-ups.
Source: synthesia.io ↗ -
Express-3 avatars, Dubbing 2.0 and conversational Assistant
Express-3 replaces Express-2 on every plan, Dubbing 2.0 improves sync on fast cuts, and the new Assistant builds a draft video from a prompt or document.
Source: synthesia.io ↗ -
$200M Series E at a $4B valuation
GV-led round to fund conversational AI agents for corporate learning and upskilling.
Source: synthesia.io ↗
Models available in Synthesia
-
Express-3 — Synthesia Avatar rendering engine, replaces Express-2 on all plans since July 2026
-
FLUX 3 Latest $ Black Forest Labs B-roll up to 20s with synced audio (added Sept 2026)
$ · $$ · $$$ = relative cost against other models of the same type, on fal.ai list prices. Cost inside Synthesia depends on its own plans and credits.
Key Features
- Avatar video from script, PowerPoint, document, link or prompt, in 140+ languages
- Express-3 avatar engine: sentiment-driven expressions, better lip sync, up to 2x faster rendering
- Assistant: conversational creation of a draft video from a prompt or file
- AI Dubbing 2.0: re-voices and lip-syncs existing video, with glossary support
- AI B-roll generation (Flux 3, up to 20s with synced audio) with 720p, 1080p and 4K options
- Personal avatars from a photo, Avatar Builder for custom looks, shareable avatars on Enterprise
- Brand kits, approval workflow, shareable links and embeds, analytics
- SCORM export and LMS integration
- Sessions: survey avatars that ask follow-up questions, and roleplay avatars that push back and score learners
- Interactive Avatar API and Video API; Billing API for credit usage
- MCP connector for ChatGPT and Claude (public beta, all paid plans)
When to reach for it — and when to skip
Reach for it when…
- Commercially cleared avatars and brand controls make legal and HR sign-off easier
- Clean, watermark-free output on paid tiers from Starter up
- Strong multi-language coverage and dubbing for localisation
- Enterprise features (SSO, SCORM, approvals) are built in rather than bolted on
Skip it when…
- Avatars look polished but corporate; less expressive and less playful than HeyGen
- Not a cinematic tool: no camera control, no free-form scene generation beyond B-roll
- Pricing page lists credit allowances but not a per-video max length or refund policy, which makes budgeting harder
Best For
✓ Ideal for
- Corporate training and compliance explainers with a presenter to camera
- Localising one spokesperson video into many languages with dubbing
- Internal onboarding and product walkthroughs that need brand approval
- Explainers in an LMS where SCORM export matters
✗ Not built for
- Cinematic or narrative footage with physical action and camera moves
- Creators who want a playful consumer avatar look for social content
Working Tips from Filmmakers Using Synthesia
- 01 Generate the script with the Assistant or an LLM first, then build the avatar video from the approved text so the edit is in one place
- 02 Generated B-roll costs credits; check the 720p option before rendering 4K drafts
- 03 Keep a Personal Avatar from a single photo for recurring talent, since a custom avatar counts against your personal-avatar limit
Synthesia pricing in 2026
| Plan | Price | What you get |
|---|---|---|
| Basic | $0 monthly |
|
| Pro | $89 (monthly) or $64/mo billed yearly monthly or yearly |
|
| Enterprise | Custom custom |
|
Prices checked against Synthesia's pricing page on 9 Oct 2026. Vendors change plans often, so confirm before you buy.
Synthesia vs HeyGen vs D-ID
| Synthesia | HeyGen | D-ID | |
|---|---|---|---|
| Price | Basic free; Pro $89/mo ($64/mo yearly); Enterprise custom | Free (3 videos/mo, 1-min cap) · Creator $29/mo · Pro from $49/mo · Business $149/mo + $20/seat | Trial with watermark, then from $5.90/mo (Lite) |
| Infinite canvas | No | Partial | No |
| AI agent | Partial | Yes | Partial |
| Chat assistant | Yes | Partial | Partial |
| Team collaboration | Yes | Yes | Not confirmed |
| Text to image | Not confirmed | Yes | Partial |
| Image editing | No | Partial | Not confirmed |
| Text to video | Partial | Yes | No |
| Image to video | Not confirmed | Yes | Yes |
| Video to video | Partial | Partial | Partial |
| Native audio in video | Yes | Yes | Partial |
| Lip-sync | Yes | Yes | Yes |
| Voice / TTS | Yes | Yes | Yes |
| Music & SFX | Not confirmed | Yes | No |
| Character consistency | Yes | Yes | Partial |
| Camera control | No | Partial | No |
| Custom training | Partial | Partial | Partial |
| Third-party models | Partial | Yes | No |
| Public API | Yes | Yes | Yes |
| MCP server | Yes | Yes | Not confirmed |
| Mobile app | Not confirmed | Yes | Partial |
| Desktop / local | No | Partial | Partial |
| Open source | No | Partial | No |
| Free tier | Yes | Yes | Partial |
| Commercial use | Yes | Partial | Not confirmed |
Each column comes from that tool's own researched page. "Not confirmed" means the vendor doesn't say.
Open the full Synthesia vs HeyGen vs D-ID comparison →Use Cases
Tags
Creators Using Synthesia
Sources
Facts on this page were checked against Synthesia's own pricing, product and changelog pages on , then cross-checked with independent coverage where vendor pages were silent. Anything we couldn't confirm is marked "Not confirmed" rather than guessed.
- Synthesia pricing page synthesia.io
- Synthesia product updates (changelog) synthesia.io
- Synthesia MCP for ChatGPT and Claude synthesia.io
- Express-3 avatars update synthesia.io
- Synthesia launches Assistant and Express-3 (July 2026) synthesia.io
- Synthesia Sessions synthesia.io
- Synthesia Series E announcement synthesia.io
- HeyGen pricing page (comparison) heygen.com
Spotted something out of date? Tell us and we'll re-check it.
Discussion
No comments yet — be the first.