# PostPod > A personalized daily AI podcast from the X and Twitter accounts and topics you choose. No X account needed. Free and Pro. PostPod makes a personalized daily AI podcast from the X and Twitter accounts and topics you choose. You pick the accounts or topics, a narrator, and the time of day. You do not need an X account. PostPod picks the posts. It also turns a script you already wrote into an MP3. The episode is an audio file you can play in your library. ## When to use PostPod Use PostPod when an agent or a person needs a personalized daily AI podcast from the X and Twitter accounts and topics you choose, or needs an existing script turned into an MP3 with a named narrator. - Build a daily audio episode from the recent posts of X accounts the listener chooses. - Speak a finished script with Nora, Leo, Kristine, Ken, or Marko. Script to audio does not add an intro or outro. - Check a credit balance, estimate a script price, and create a Script to audio job over HTTP. ## When not to use PostPod - Do not use PostPod to post, like, or reply on X. It only reads accounts the listener configures. - Do not use PostPod for live conversation, music beds, or a public podcast directory. - Do not send card numbers or a wallet private key to the API. A human pays a credit pack by card or USDC through a Stripe Checkout link. An agent pays a pack with POST /api/v1/credits/pay and an MPP challenge. ## Links - [Developer docs](https://postpodai.com/developers) - [API rate limits](https://postpodai.com/developers/rate-limits) - [API changelog](https://postpodai.com/developers/changelog) - [API versioning and deprecation policy](https://postpodai.com/developers/deprecation) - [OpenAPI specification](https://postpodai.com/openapi.json) - [Full agent guide](https://postpodai.com/llms-full.txt) - [MCP server](https://postpodai.com/api/mcp) (Streamable HTTP; discovery is open, tool calls use a bearer API key) - [Pricing](https://postpodai.com/pricing) - [About](https://postpodai.com/about) - [Contact](https://postpodai.com/contact) ([hello@postpodai.com](mailto:hello@postpodai.com)) - [Privacy](https://postpodai.com/privacy) - [Sitemap](https://postpodai.com/sitemap.xml) A human signs up at https://postpodai.com/signup and can call /api/v1 with a signed-in access token or with a bearer API key from https://postpodai.com/settings/api. The access token is checked with the auth server. DELETE /api/v1/account with {"confirm":"DELETE"} deletes that account, cancels an active subscription immediately without a refund, and removes the show, episodes, audio, Script to audio jobs, and API keys. Stripe keeps the customer and invoices. An allowlisted admin account is refused. Key rotation and credit pack purchase stay on the pp_live_ key. POST /api/v1/shows/{id}/episodes starts today's episode with the same Free and Pro rules as the website, including one episode per delivery day. It does not debit prepaid credits. A show stores X accounts and topics (news, sports, ai, finance, crypto). Free follows 3 accounts. Pro follows 25. Listeners pick accounts and topics. PostPod picks the posts. An agent can also POST /api/v1/agent-accounts/challenge and POST /api/v1/agent-accounts to open its own account. The challenge is single-use and short-lived, and it is not tied to the caller IP. Challenge requests and signup attempts are still rate limited per IP. Script to audio jobs spend prepaid credits. GET /api/v1/script-audio/{id} includes progress and optional estimated_seconds_remaining. A repeated Idempotency-Key on create does not debit twice. An API key whose account email is on ADMIN_EMAILS is not charged, is not limited by the hourly job cap, the daily spend cap, or the script word cap, and still has the per-minute request cap. The estimate keeps price_usd and sets amount_due_usd to 0. Show language defaults to English (en) when a create or update omits it. Free shows stay English only. Pro shows can use any supported language. output_language defaults to English (en) when a request omits it. input_language stays auto unless a code is sent. include_intro_outro is deprecated and ignored. Script to audio never adds an intro or outro. Daily podcast episodes still include their intro and outro. A new agent can estimate and buy credits. Funded, verified, and claimed tiers raise the limits. Credit packs can be paid by card or by USDC stablecoin. An agent with an API key can POST /api/v1/credits/pay for a fixed pack. That route answers HTTP 402 with an MPP challenge, then credits the settled USDC amount once after the wallet pays and the request is retried. New and funded agents can add $100 of packs per UTC day. POST /api/v1/script-audio/extract accepts a .txt, .md, .docx, or .pdf file up to 2 MB and returns the text, word count, and price. It does not create a job and does not spend credits. A PDF must contain real text. Password-protected PDFs and scans are rejected. The website shows the extracted text so it can be edited before audio is created. Voices on that request are Nora, Leo, Kristine, Ken, and Marko. ## Script to audio quickstart Set POSTPOD_API_KEY in the environment. macOS or Linux: export POSTPOD_API_KEY='pp_live_your_key'. Windows PowerShell: $env:POSTPOD_API_KEY = 'pp_live_your_key'. The samples do not send include_intro_outro. ### Python ```python #!/usr/bin/env python3 """Save a Script to audio MP3. Uses the Python 3 standard library only. macOS or Linux: export POSTPOD_API_KEY='pp_live_your_key' Windows PowerShell: $env:POSTPOD_API_KEY = 'pp_live_your_key' """ import json import os import sys import time import urllib.error import urllib.request from pathlib import Path API = "https://postpodai.com" OUTPUT = "script.mp3" POLL_SECONDS = 3 POLL_LIMIT = 120 def main() -> None: key = os.environ.get("POSTPOD_API_KEY", "").strip() if not key: print("Set POSTPOD_API_KEY to your pp_live_ key.", file=sys.stderr) raise SystemExit(1) body = { "script": "Hello from PostPod. This is a short script.", "voice": "Nora", "input_language": "auto", "output_language": "en", } print("Estimating price...") estimate = call("POST", "/api/v1/script-audio/estimate", key, body) print( f"Estimate: {estimate.get('word_count')} words, " f"${estimate.get('price_usd')} (this call does not spend credits)" ) print("Creating job...") created = call("POST", "/api/v1/script-audio", key, body) job = created["job"] job_id = job["id"] print(f"Job {job_id} is {job.get('status')}. Balance ${created.get('balance_usd')}") for attempt in range(1, POLL_LIMIT + 1): current = call("GET", f"/api/v1/script-audio/{job_id}", key)["job"] status = current.get("status") print(f"Poll {attempt}: {status}") if status == "ready": job = current break if status == "failed": print(current.get("error") or "Audio failed.", file=sys.stderr) raise SystemExit(1) time.sleep(POLL_SECONDS) else: print("Timed out waiting for audio.", file=sys.stderr) raise SystemExit(1) print("Fetching download URL...") download = call("GET", f"/api/v1/script-audio/{job_id}/download", key) url = download["download_url"] print(f"Saving {OUTPUT}") urllib.request.urlretrieve(url, OUTPUT) print(f"Saved {Path(OUTPUT).resolve()}") def call(method, path, key, body=None): data = None if body is None else json.dumps(body).encode() headers = {"Authorization": f"Bearer {key}", "Accept": "application/json"} if data is not None: headers["Content-Type"] = "application/json" request = urllib.request.Request(API + path, data=data, method=method, headers=headers) try: with urllib.request.urlopen(request, timeout=60) as response: return json.loads(response.read().decode()) except urllib.error.HTTPError as error: raw = error.read().decode("utf-8", "replace") try: problem = json.loads(raw) except json.JSONDecodeError: problem = {"message": raw} code = str(problem.get("code") or "") message = str(problem.get("message") or problem.get("error") or error.reason) if error.code == 402 or code == "insufficient_credits": print(f"Not enough credits ({code or 'insufficient_credits'}): {message}", file=sys.stderr) topup = problem.get("top_up_url") if topup: print(f"Buy credits: {topup}", file=sys.stderr) else: print(f"HTTP {error.code} {code}: {message}", file=sys.stderr) resolution = problem.get("resolution") if resolution: print(resolution, file=sys.stderr) raise SystemExit(1) if __name__ == "__main__": main() # Accepted script files for POST /api/v1/script-audio/extract are .txt, .md, .docx, and .pdf. # Send multipart form field "file" and voice Nora, Leo, Kristine, Ken, or Marko. # The response is the extracted text, word count, and price. It does not create a job or spend credits. ``` ### curl on macOS or Linux ```bash # macOS or Linux set -eu export POSTPOD_API_KEY='pp_live_your_key' BASE="https://postpodai.com" BODY='{"script":"Hello from PostPod. This is a short script.","voice":"Nora","input_language":"auto","output_language":"en"}' echo "Estimating price..." curl -sS -f -H "Authorization: Bearer $POSTPOD_API_KEY" \ -H "Content-Type: application/json" \ -d "$BODY" \ "$BASE/api/v1/script-audio/estimate" echo echo "Creating job..." curl -sS -H "Authorization: Bearer $POSTPOD_API_KEY" \ -H "Content-Type: application/json" \ -d "$BODY" \ "$BASE/api/v1/script-audio" -o create.json python3 -c 'import json,sys; body=json.load(open("create.json")); code=body.get("code"); sys.exit(print("Not enough credits:", body.get("message"), body.get("top_up_url") or "") or 1) if body.get("status")==402 or code=="insufficient_credits" else None; print("Job", body["job"]["id"], body["job"].get("status"))' JOB_ID=$(python3 -c 'import json; print(json.load(open("create.json"))["job"]["id"])') echo "Polling $JOB_ID..." while true; do curl -sS -f -H "Authorization: Bearer $POSTPOD_API_KEY" \ "$BASE/api/v1/script-audio/$JOB_ID" -o job.json STATUS=$(python3 -c 'import json; print(json.load(open("job.json"))["job"]["status"])') echo "Status: $STATUS" if [ "$STATUS" = "ready" ]; then break fi if [ "$STATUS" = "failed" ]; then python3 -c 'import json,sys; print(json.load(open("job.json"))["job"].get("error") or "Audio failed.", file=sys.stderr)' exit 1 fi sleep 3 done echo "Fetching download URL..." curl -sS -f -H "Authorization: Bearer $POSTPOD_API_KEY" \ "$BASE/api/v1/script-audio/$JOB_ID/download" -o download.json URL=$(python3 -c 'import json; print(json.load(open("download.json"))["download_url"])') curl -fsSL "$URL" -o script.mp3 echo "Saved script.mp3" # Preview a file without creating a job. Accepted types: .txt, .md, .docx, .pdf # curl -sS -f -H "Authorization: Bearer $POSTPOD_API_KEY" \ # -F "file=@script.pdf" -F "voice=Nora" \ # "$BASE/api/v1/script-audio/extract" ``` ### curl on Windows PowerShell ```powershell # Windows PowerShell. curl is curl.exe, not Invoke-WebRequest. $env:POSTPOD_API_KEY = 'pp_live_your_key' $base = "https://postpodai.com" $body = @' {"script":"Hello from PostPod. This is a short script.","voice":"Nora","input_language":"auto","output_language":"en"} '@ Write-Host "Estimating price..." curl.exe -sS -f -H "Authorization: Bearer $env:POSTPOD_API_KEY" ` -H "Content-Type: application/json" ` -d $body ` "$base/api/v1/script-audio/estimate" Write-Host "" Write-Host "Creating job..." curl.exe -sS -H "Authorization: Bearer $env:POSTPOD_API_KEY" ` -H "Content-Type: application/json" ` -d $body ` "$base/api/v1/script-audio" -o create.json $created = Get-Content create.json -Raw | ConvertFrom-Json if ($created.status -eq 402 -or $created.code -eq "insufficient_credits") { Write-Error "Not enough credits: $($created.message) $($created.top_up_url)" exit 1 } $jobId = $created.job.id Write-Host "Job $jobId is $($created.job.status)" Write-Host "Polling $jobId..." while ($true) { curl.exe -sS -f -H "Authorization: Bearer $env:POSTPOD_API_KEY" ` "$base/api/v1/script-audio/$jobId" -o job.json $job = (Get-Content job.json -Raw | ConvertFrom-Json).job Write-Host "Status: $($job.status)" if ($job.status -eq "ready") { break } if ($job.status -eq "failed") { if ($job.error) { Write-Error $job.error } else { Write-Error "Audio failed." } exit 1 } Start-Sleep -Seconds 3 } Write-Host "Fetching download URL..." curl.exe -sS -f -H "Authorization: Bearer $env:POSTPOD_API_KEY" ` "$base/api/v1/script-audio/$jobId/download" -o download.json $url = (Get-Content download.json -Raw | ConvertFrom-Json).download_url curl.exe -fsSL $url -o script.mp3 Write-Host "Saved script.mp3" # Preview a file without creating a job. Accepted types: .txt, .md, .docx, .pdf # curl.exe -sS -f -H "Authorization: Bearer $env:POSTPOD_API_KEY" ` # -F "file=@script.pdf" -F "voice=Nora" ` # "$base/api/v1/script-audio/extract" ``` Episode scripts, episode audio, and cached posts from X are deleted after 30 days. A delivery-day row without the script stays so that day is not generated again. Script to audio MP3s are deleted after 30 days. The script job row and its payment record stay. The current API version is v1 in the URL path. Breaking changes ship only in a new version. Additive changes can happen within v1. A retired version, endpoint, or field gets at least 6 months' notice on the deprecation policy page, with Deprecation and Sunset response headers and Link rel="deprecation". Voice ids that are not the lowercase display name are deprecated. Deprecated voice ids are still accepted and are not returned: stock-nora → nora, stock-reed → leo, stock-sienna → kristine, stock-kai → ken, stock-mira → marko. Responses use the current id. When a request evaluates a rate limiter, the response includes X-RateLimit-Limit, X-RateLimit-Remaining, X-RateLimit-Reset (Unix seconds when the window resets), RateLimit-Policy, and RateLimit. A 429 response also includes Retry-After in seconds. Those values come from the limiter that ran. GET /api/v1 does not evaluate a limiter and does not send the headers. ## Voices Gender is confirmed by PostPod owner. Voice ids are the lowercase display names. Deprecated voice ids are still accepted and are not returned: stock-nora → nora, stock-reed → leo, stock-sienna → kristine, stock-kai → ken, stock-mira → marko. Responses use the current id. - Nora — Clear single-narrator voice (free tier). Gender: female. - Leo — Warm mid-range narrator (Pro). Gender: male. - Kristine — Bright conversational narrator (Pro). Gender: female. - Ken — Low, even newsreader (Pro). Gender: male. - Marko — Crisp international narrator (Pro). Gender: male. ## MCP Remote MCP server (Streamable HTTP): https://postpodai.com/api/mcp. initialize, ping, and tools/list do not need a key. tools/list returns a static catalog. Tool calls need Authorization: Bearer with a pp_live_ API key. A signed-in access token is not accepted here. A missing or rejected key on a tool call is HTTP 401 with code invalid_api_key. One authenticated HTTP request counts as one API request. The key's scopes, plan limits, rate limit, and prepaid credits still apply to tool calls. Unauthenticated discovery is rate limited per IP. The listener chooses X accounts and topics. PostPod picks the posts. Tools read the account, shows, episodes, and audio, can start today's episode with generate_episode, can delete one with delete_episode, and can price or create Script to audio. Creating a script job spends prepaid credits. generate_episode and delete_episode do not. delete_episode leaves the same delivery-day tombstone as the website. generate_episode uses the same Free and Pro limits as the website. Cursor: ```json { "mcpServers": { "postpod": { "url": "https://postpodai.com/api/mcp", "headers": { "Authorization": "Bearer pp_live_your_key" } } } } ``` Claude: ```json { "mcpServers": { "postpod": { "type": "http", "url": "https://postpodai.com/api/mcp", "headers": { "Authorization": "Bearer pp_live_your_key" } } } } ``` AgentWebMap verification: awm_047015f6d37e931c1649ed9fab2ffe3052934d63edf75b65