Monday through Friday
DALL-E killer?
This newsletter was brought to you byDigitalOceanJoin us for our next event
The Product Hunt team is hosting a meetup in Berlin (cohosted by FastGen) on January 8th, and we'd love for you to kick back and demo products with us! RSVP here.
Here’s some news:
🐁 Micky Mouse, or a version of him, is in the public domain, and chaos has ensued.
👓 Want to try VR? Meta has cut the price for its Quest 2 headset and accessories.
💲 Bitcoin is on the move, rising to above $45k for the first time since 2022.
Google just dropped a new video-based AI
New year, new AI. The year 2023 saw some massive launches in the world of AI with the likes of GPT-4, Gemini, and Midjourney V6, to name a few, and it looks like 2024 will be no different.
Google is coming out swinging in the new year with its latest model: VideoPoet. VideoPoet is the latest addition to Google’s promised “multimodal AI,” meaning it can generate text, audio, images, and even video. The last part is where VideoPoet fits in.
VideoPoet is an experimental large language model (LLM) that’s capable of a number of different video-related generation tasks, including text-to-video, image-to-video, and even video-to-audio.
Take text-to-video as an example. You can input something as ridiculous as a “raccoon sipping wine in Times Square,” and within a few seconds, you will have the generated video ready to be shared in whatever style you requested. Just imagine the Mona Lisa lip-syncing an ABBA song.
Already have an image you want animated? Drop it in, and VideoPoet will generate a short clip based on the image. Once the video is generated, you can edit it further and stylize it to how you want just by telling the AI your instructions.

Inference should feel simple. Once AI hits production traffic, you're really managing routing, retries, observability, GPU bills running while nobody's home.
DigitalOcean Serverless Inference handles all of that in one API — OpenAI- and Anthropic-compatible, 55+ models including Claude, DeepSeek, Qwen, and Llama. Scale to zero when you're not using it. Pay per token when you are. VPC and zero data retention on by default.
230 tokens/sec on DeepSeek V3.2 — 3.9× faster than AWS Bedrock (Artificial Analysis). Runs on the same bill as your databases and Kubernetes. One less thing to reconcile.
Monday through Friday
Our ultra-fast Daily: Three takes on new products. Yesterday’s top ten launches. That’s it.

