REST API

Schedule posts from a DeepSeek agent

If you are running DeepSeek for cost reasons, the posting layer should not undo that. One API call schedules to every network you have connected.

Start for free Full docs

Setup

Connect DeepSeek in three steps

Create an API key

In SchedPilot, open API Access and generate a key. It starts with smm_.

Give your agent the endpoint

Point it at the developer API with your key as a bearer token.

Let it schedule

Posts land in your calendar for approval, exactly as they would from the app.

curl -X POST https://api.schedpilot.com/developers/v1/post \
  -H "Authorization: Bearer smm_your_key" \
  -H "Content-Type: application/json" \
  -d '{"content":"Shipping today","date":"2026-07-10","time":"09:00",
       "accounts":[{"id":12,"type":"linkedin"}]}'

Replace smm_your_key with the key from API Access in your SchedPilot account.

Once connected

What DeepSeek can do

get_accounts

List every connected social account, with its id and type.

create_post

Schedule a post to one or many accounts, with media, a thread and a video title.

list_posts

Read scheduled and published posts, filtered by date or status.

delete_post

Cancel a scheduled post before it goes out.

upload_media

Upload an image or video from a URL and get a media id back.

get_analytics

Read impressions, reach, likes, comments and shares for a published post.

Rate limits are 150 read and 100 write requests per hour, per user. Scheduling a post with media costs two writes.

Why DeepSeek

Cheap to run, still reaches everywhere

Teams pick DeepSeek to keep inference costs down. It would be odd to then pay per network for distribution — one flat plan and one endpoint keeps the economics intact.

A single endpoint for all nine networks
No per-network pricing
Self-hosted or API-hosted models both work

Try asking it

Things that work well once DeepSeek is connected.

“Generate a week of posts and schedule them”

“Queue this draft for tomorrow at 9am”

“List scheduled posts for review”

We have not verified MCP support, so this uses the REST API — which works equally well with a self-hosted model.

FAQs

DeepSeek questions

Does a self-hosted model work?

Yes. All that matters is that something in your stack can make an HTTPS request.

What is the rate limit?

150 read and 100 write requests per hour, per user.

Can it publish without my approval?

No. Scheduling queues a post; it waits in your calendar until you approve it. That holds for every agent and every connection method.

Using something else?

There is a setup guide for every agent we support, plus a REST API for anything that is not on the list.

See all agents

Connect DeepSeek to SchedPilot

Free to start, and every post waits for your approval.

Start for free