Blog
Sume blog: agent-native video generation, Avatar 1.0 orchestration, and production APIs for builders and coding agents.
aiohttp 3.14.4 fixes truncated retried file bodies: Sume submits
aiohttp 3.14.4 fixed an internal retry that sent a cut-off body from a file object. Send a Sume JSON submit from bytes and keep one Idempotency-Key.
Express 5.3 logs Error.cause: keep the Sume request_id
Express 5.3.0 prints the full error object, nested cause included. Wrap a failed Sume call so its code and request_id land in the log line you grep.
FastAPI 0.143 makes OTel exporter setup opt-in: Sume webhook spans
FastAPI 0.143.0 stops auto-configuring OpenTelemetry exporters. If your Sume webhook route traces went quiet after the upgrade, here is the one-line opt-in.
Go 1.27.2 and 1.26.9 patch net/http: update your Sume webhook receiver
Go 1.27.2 and 1.26.9 (2026-10-08) list security fixes in net/http and crypto/tls. A stdlib receiver for Sume webhooks that verifies the signature, in two files.
Node 26.11 --process-timeout exits 124: the Sume job keeps running
Node 26.11.0 adds --process-timeout, which kills a script with exit code 124. A Sume job it was waiting on keeps running and billing, so save the polling URL.
Vercel skipMiddlewareRequestBody and a Sume webhook route
Vercel's Oct 8 skipMiddlewareRequestBody stops sending request bodies to Routing Middleware. What it means for a Sume webhook route that signs the raw body.
Suno Speech makes voice and music as one track. Sume keeps them apart
Suno Speech beta renders voice and music together. On Sume you generate speech, a music bed and a ducked mix as separate jobs, so each can be redone alone.
Bun 1.4.3 ERR_PROXY_TUNNEL: is it a Sume error or your proxy?
Bun 1.4.3 rejects fetch with ERR_PROXY_TUNNEL on a failed CONNECT. A Sume error always carries error.request_id; use it to tell the two apart.
Bun 1.4.3 fake timers: test a Sume poll loop with request timeouts
Bun 1.4.3 fixes advanceTimersByTime spinning with AbortSignal.timeout. Test a Sume job poll loop that honors next_poll_after_seconds without real waits.
Bun 1.4.3 fetch Content-Length check: send a Sume JSON body as bytes
Bun 1.4.3 rejects fetch when a declared Content-Length mismatches a stream body. For a Sume submit, skip the header and pass the encoded JSON.
Gift box insert: a QR code to a setup video you can swap later
Put a QR code on a gift-box insert that opens a setup video. Make the clip on Sume, caption it for sound-off, draw the QR in code, and point it at a redirect.
Lemon8 video length and ratio: 60 s, 3:4, prepared on Sume
Lemon8 supports videos up to 60 seconds and favors 3:4. Trim a longer clip to 60 s, crop 16:9 to 3:4 and conform to 1080x1440 with Sume.
Referral program launch video: avatar above the share screen
Explain a refer-a-friend offer in 30 seconds: an avatar script with a silent beat, the share screen stacked below it, and a plan check before you pay.
TikTok pop-out showcase: 8 s minimum, 3 s in, 5 s tail
TikTok's pop-out add-on needs video of at least 8 s; the pop-out shows 3 s in and ends 5 s before the end. Work out the window and trim a clip to fit on Sume.
Which audiobook stores accept AI narration? ACX, Apple, KDP, Spotify
ACX bans unauthorized AI narration, Apple narrates for you, KDP is an invite beta, Spotify wants a disclosure. Read 2026-10-10, plus where Sume TTS audio fits.
Auphonic audio cleanup vs Sume audio detach: different jobs
Auphonic cleans and levels audio; Sume audio detach only pulls the track out of a video as wav or mp3. Compare billing, limits and where each belongs.
Claude Code hook: swap Sume's download URL for a saved file
A PostToolUse hook with updatedToolOutput can fetch the signed link from assets_download_url, save the file, and show Claude only the local path.
Copilot Studio hooks fail open: cap paid media spend server-side
Copilot Studio hooks can block a tool call, but a failed hook lets the agent continue. Put the media spend limit where it can't fail open: the Sume run cap.
Cursor cloud agents and Sume MCP: HTTP, redacted headers
Cursor cloud agents run MCP over HTTP or stdio. Sume's hosted server is HTTP, so the key stays in Cursor's backend and OAuth is per user. What to set.
Elai 75+ languages vs Sume Avatar 1.0 English-only
Presenter video in many languages? Elai lists 75+ languages and 80+ avatars; Sume Avatar 1.0 is English-only. What fits, and what Sume still does around it.
Clinic flu-shot reminder video: four 20-second avatar clips
Make a flu-shot reminder video for a clinic or pharmacy with a Sume avatar: what four 20-second clips cost at each quality tier, plus captions.
Gemini agent keeps memory for days; Sume API runs start fresh
Google's Gemini agent keeps one memory across devices and days. Sume's API runs start in a new thread, so state travels as input, files and previous_run_id.
Gemini CLI replaced by Antigravity CLI: move your Sume MCP entry
Google says Antigravity CLI replaced Gemini CLI for unpaid and Google One users. Move Sume's MCP entry from settings.json httpUrl to mcp_config.json serverUrl.
Gemini Skills replace Gems: a saved workflow you call by API
Google is replacing Gems with Skills, reusable instructions in the Gemini app. For a saved video recipe your backend can call, Sume Formats take a POST.
Gladia diarization vs Sume speech-to-text: what you can switch
Gladia lists speaker diarization and word timestamps on every plan. Sume STT always returns word timings but fixes diarization server-side, with no switch.
Google Discover and Article schema image sizes (16:9, 4:3, 1:1)
Google Discover wants 1200 px wide, 16:9 images; Article schema wants 16:9, 4:3 and 1:1. Here is how to generate all three ratios through the Sume image API.
Google Vids point-in-time editing: timed text via Sume cues
Google Vids now shows only what is on screen at the playhead. Sume has no canvas, but caption cues with start and end place timed text by API.
n8n Gemini node adds Nano Banana 2.1: use it or call Sume?
n8n 2.43.3 lists Nano Banana 2.1 and 2 Lite in the Gemini node. Sume lists 2.1 but no Lite id; here is when to use the node and when an HTTP Request fits.
Nano Banana 2.1 small text looks blurry? A three-row test on Sume
Google's Nano Banana 2.1 card says small text is often blurry. Run one poster prompt on three Sume rows for $5.40 for 12 prompts and pick by eye.
Nano Banana 2.1 replies with text and images; Sume returns URLs only
Google says Nano Banana 2.1 outputs text as well as images. Sume's image response lists image URLs and usage, so write captions and alt text in a separate step.
Narakeet Markdown-to-video API vs Sume Timeline render
Narakeet turns a Markdown script into a narrated video; Sume Timeline assembles audio you already have with clips. Limits, rates and how to combine them.
nginx proxy_read_timeout is 60s, but Sume gives up at 10s
nginx waits up to 60 seconds for your app by default, yet a Sume webhook attempt times out at 10. Why raising proxy timeouts does not help and what to do.
NVIDIA Magpie TTS for voice agents vs Sume TTS 1.0: which to use?
Magpie TTS is an open-weights, low-latency model for live voice agents in 12 languages. Sume TTS is async and non-streaming. How to choose, read 2026-10-10.
Page on-call when a Sume webhook is exhausted: PagerDuty dedup_key
Send a PagerDuty Events API v2 trigger when Sume gives up on a webhook, using the job id as dedup_key so retries and redelivers never page twice.
NVIDIA Parakeet v3 or Canary v2 vs a hosted STT API: Sume STT 1.0
Parakeet TDT v3 and Canary 1B v2 are open ASR models for 25 European languages. When to self-host them and when Sume STT at $0.01 a minute is enough.
Sonix SRT and VTT export vs Sume burned-in captions
Need a subtitle file? Sonix exports SRT and VTT; Sume video captions return a captioned video, not a sidecar file. Compare outputs, languages and $0.20 jobs.
Sony Woosh sound effects model: does Sume have SFX or video-to-audio?
Sony AI's Woosh makes sound effects and video-to-audio. Sume has no SFX route; it offers video models with native audio, plus music stings. What to use instead.
Suno Speech makes voice and music in one pass. What does Sume do?
Suno's Speech beta (Oct 1, 2026) makes voice and music as one track. Sume has no such model; it layers TTS, Music and a Timeline soundtrack. Trade-offs inside.
Terraform and secrets: keep the Sume API key out of state
Terraform's sensitive flag hides a value from logs but still writes it to state. How to handle a Sume API key and webhook signing secret without that leak.
Compose 400: stack and overlay layout keys cannot be mixed
Fix compose_stack_takes_no_overlay_layout and its mirror: pick stack or overlay as operation, then send only the layout keys that belong to it.
Timeline 400: first slot must start at 0 and take no transition
Fix timeline_must_start_at_zero and transition_on_first_segment: start video[0] at 0, use source_in to skip, put fades on slot two, then plan for free.
Upscayl free desktop 16x vs Sume image upscale API at 4x
Upscayl is a free open-source desktop upscaler; Sume's image upscale is a $0.20 API call capped at a 4x factor. How to choose for a batch of photos.
video_inspect 400: frames program and transcribe flag errors
Fix video_inspect_frames_program_conflict, _required and video_inspect_transcribe_required: send at or fps, and transcribe true with its language options.
Voicemaker 3,000-character cap vs a 20,000-character Sume TTS call
Voicemaker caps each conversion at 3,000 to 10,000 characters by plan; Sume TTS accepts up to 20,000 characters per request. What that means for long scripts.
Context stitching and streaming TTS: Eleven v4 vs Sume async jobs
Eleven v4 lists context stitching and bidirectional streaming. Sume TTS is async and non-streaming, so keep a section in one job and join sections.
Laugh and sigh tags in TTS: Gemini 3.8 tags vs Sume's emotion field
Gemini 3.8 TTS takes inline tags such as laugh, sigh and breath. Sume TTS documents an emotion hint, speed and volume instead; test tags on one short job.
How many words fit in a 30-second voiceover, and what it costs
A 30-second voiceover is about 75 words or 450 characters at a planning pace of 150 wpm. Sume TTS bills characters, so here is the length-to-cost table.
IVR phone menu prompts with TTS: 8 kHz mu-law WAV and cost per prompt
Make phone-menu prompts as 8 kHz mu-law or A-law WAV with Sume TTS. A 20-prompt menu bills 20 cents because each short job hits the 1-cent minimum.