Skip to main content
POST
Send a message and start a generation
This returns immediately with status: "generating" — it does not return the finished video. Poll GET /fragments/{fragmentId} until animationUrl is non-null.

Authorizations

x-api-key
string
header
required

Your Mathify API key (mk_live_<64 hex chars>), generated at /settings/api-keys.

Path Parameters

projectId
string
required

Body

application/json
content
string
required
Required string length: 1 - 100000
Example:

"Animate the Pythagorean theorem"

model
enum<string>

Optional preference for the model or provider family used by the main generation agent — the one doing the primary reasoning and code generation. Other steps in the pipeline (e.g. scene planning, visual validation) may already use a different model regardless of this setting. Even for the main agent, it's a preference rather than a guarantee: depending on the difficulty or type of your prompt, Mathify may route generation to a different model or provider family if we judge it a better fit for that specific task. Omit this field to let Mathify choose automatically.

Available options:
openai,
anthropic,
qwen,
gemini,
grok,
gptoss
Example:

"anthropic"

forceWatermark
boolean

Opt into a watermarked render even if the caller's account has the no-watermark benefit.

Response

Generation started

messageId
string
required
fragmentId
string
required
status
string
required
Allowed value: "generating"
totalSteps
integer
required
Required range: -9007199254740991 <= x <= 9007199254740991