Tech Stacks · AI
OpenAI
Category
AI
At the studio
Since 2022
Projects shipped
Numerous
Status
Active
(01) Our take
OpenAI remains the most common LLM provider in production across the industry, and we support it in any engagement that requires it — whether because a client already runs on it, because the specific use case favours it, or because cost at scale tips the calculus.
We default to Claude for quality and reasoning tasks, but OpenAI's embedding models, Whisper, and DALL-E remain best-in-class for their respective domains. Most mature AI engagements we ship use multiple providers through a single abstraction layer.
(02) What we build with it
Typical work
- Embedding generation (text-embedding-3)
- Speech-to-text via Whisper
- Image generation and analysis via DALL-E and GPT-4V
- GPT-powered features for teams already invested in OpenAI
(03) How we engage
Engagement shape
Rarely standalone — usually part of a larger AI feature build through the Vercel AI SDK or similar abstraction.
(04) Pairs with
Works well with
Technologies we reach for alongside OpenAI in most engagements.
Thinking about a OpenAI project?