Syren is Synthesia's new AI video agent that turns a chat prompt into a finished, agency-style video and then lets you refine it conversationally instead of rebuilding scenes by hand. It targets marketing, product, and enablement teams who need launch videos, explainers, how-tos, and ads without hiring a production crew, and this launch opens it up as a free browser experience with a Claude MCP integration for teams that already work inside Claude.
Under the hood, Syren pairs Opus 5.5 with Synthesia's own renderer, which adds 3D support alongside motion graphics, voiceover, music, and avatars. The agent builds the video while you watch, writing it as code and showing a live preview as it works, and you can feed it documents, images, PowerPoints, or links to ground the output in real source material. Syren also learns a brand identity from existing videos, picking up colors, typography, animation style, and editing rhythm , which is the piece that pushes output past generic AI video toward something a company can actually ship.
The release lands as part of Synthesia's broader 3.0 push beyond talking-head avatars into fully generated, interactive video. Synthesia was co-founded in 2017 by CEO Victor Riparbelli, who announced the launch, alongside a team of AI researchers , and Syren is the clearest signal yet of where the company is taking its platform next for operators evaluating the AI video category.
Comments (11)
Sign in to join the discussion.
V
Viktoras P.2d ago
That cold open where the avatar actually turns its head without the uncanny neck snap did more for me than the whole voiceover. First five seconds held, which is more than I can say for most launch videos this week.
M
Mireille Kofi2d ago
Calling it a ChatGPT moment is a bold flex to put in your own tweet. Either you ship the receipts in replies or the ratio ships you.
D
Deepa Rathore2d ago
Claude MCP support out the gate is the real headline here and you buried it under three fire emojis. Rework the hook, this deserves better billing.
S
Sanne van den Berg2d ago
Procurement question before marketing gets excited: data residency for EU renders, and does avatar training footage sit in the same region as the output. Our legal team eats products like this for breakfast.
T
Takumi Hoshi2d ago
Free to try is doing heavy lifting. What's the token math when I actually want to render a 60 second clip in 4K, because that's where every one of these tools quietly turns into a mortgage.
O
obi.exe2d ago
The 3D support demo looked crunchy in the best way. Please tell me there's a way to pin a camera path without writing a prompt that reads like a Terrence Malick screenplay.
R
Rafa Quiroga2d ago
Building in the adjacent avatar space and I'll admit the chat-to-edit loop looks tight. My only quibble is the renderer branding, Opus 5.5 and Syren in one sentence is a lot of proper nouns for a tweet.
N
Noor Haddad2d ago
Naive question from someone who just opened the browser version: where does my prompt history live and can I share a project with a teammate without emailing a link like it's 2011.
K
Kostas M.2d ago
API docs or it didn't happen. Specifically want to see rate limits, webhook payloads for render completion, and whether I can swap avatars mid-scene via a single call.
Y
Yuan Wei2d ago
Agentic video is a shrinking market dressed up as a growing one. Every enterprise buyer I talk to wants fewer talking-head explainers, not more, no matter how pretty the renderer.
H
Harpreet S.2d ago
Counterpoint: the shrinking market is bad talking-head explainers. If the chat-to-edit actually works, the ceiling is way higher than another Loom clone.
Work With Us
Book an intro call to explore our suite of services. We might have a waitlist.
Synthesia
AI Video and Avatar platform for business
AI Video and Avatar platform for business
Tags
TLVC Rating
Community Rating
No ratings yet
to rate this launch.
Syren is Synthesia's new AI video agent that turns a chat prompt into a finished, agency-style video and then lets you refine it conversationally instead of rebuilding scenes by hand. It targets marketing, product, and enablement teams who need launch videos, explainers, how-tos, and ads without hiring a production crew, and this launch opens it up as a free browser experience with a Claude MCP integration for teams that already work inside Claude. Under the hood, Syren pairs Opus 5.5 with Synthesia's own renderer, which adds 3D support alongside motion graphics, voiceover, music, and avatars. The agent builds the video while you watch, writing it as code and showing a live preview as it works, and you can feed it documents, images, PowerPoints, or links to ground the output in real source material. Syren also learns a brand identity from existing videos, picking up colors, typography, animation style, and editing rhythm , which is the piece that pushes output past generic AI video toward something a company can actually ship. The release lands as part of Synthesia's broader 3.0 push beyond talking-head avatars into fully generated, interactive video. Synthesia was co-founded in 2017 by CEO Victor Riparbelli, who announced the launch, alongside a team of AI researchers , and Syren is the clearest signal yet of where the company is taking its platform next for operators evaluating the AI video category.
Comments (11)
That cold open where the avatar actually turns its head without the uncanny neck snap did more for me than the whole voiceover. First five seconds held, which is more than I can say for most launch videos this week.
Calling it a ChatGPT moment is a bold flex to put in your own tweet. Either you ship the receipts in replies or the ratio ships you.
Claude MCP support out the gate is the real headline here and you buried it under three fire emojis. Rework the hook, this deserves better billing.
Procurement question before marketing gets excited: data residency for EU renders, and does avatar training footage sit in the same region as the output. Our legal team eats products like this for breakfast.
Free to try is doing heavy lifting. What's the token math when I actually want to render a 60 second clip in 4K, because that's where every one of these tools quietly turns into a mortgage.
The 3D support demo looked crunchy in the best way. Please tell me there's a way to pin a camera path without writing a prompt that reads like a Terrence Malick screenplay.
Building in the adjacent avatar space and I'll admit the chat-to-edit loop looks tight. My only quibble is the renderer branding, Opus 5.5 and Syren in one sentence is a lot of proper nouns for a tweet.
Naive question from someone who just opened the browser version: where does my prompt history live and can I share a project with a teammate without emailing a link like it's 2011.
API docs or it didn't happen. Specifically want to see rate limits, webhook payloads for render completion, and whether I can swap avatars mid-scene via a single call.
Agentic video is a shrinking market dressed up as a growing one. Every enterprise buyer I talk to wants fewer talking-head explainers, not more, no matter how pretty the renderer.
Counterpoint: the shrinking market is bad talking-head explainers. If the chat-to-edit actually works, the ceiling is way higher than another Loom clone.