GooseAI is an AI infrastructure service providing API access to large language models for text generation and processing, positioned explicitly as a lower-cost alternative to running the same workloads on OpenAI’s API. Available models span GPT-Neo 1.3B, Fairseq 1.3B, GPT-J 6B, Fairseq 6B, Fairseq 13B, and GPT-NeoX 20B.
“Feature parity with industry standard APIs” at roughly 30% the cost, with switching described as “as easy as changing one line of code,” is the entire value proposition – for a team already building on an OpenAI-API-shaped integration, GooseAI removes the switching cost that normally makes changing LLM providers a real engineering project rather than a configuration change. Question/answer and classification capabilities are listed as coming soon, alongside currently-available text completion/generation.
GooseAI targets developers and organizations currently on OpenAI’s API looking to cut inference costs. Pricing scales by model size: small models $0.000110/request, medium $0.000450/request, large $0.001250/request, massive $0.002650/request.
