Google AI (Gemini API / Vertex AI) review

Input fees of $0.25 to $4 and output fees of $1.50 to $18 cover the range across the catalogue. New accounts also get $300 in free credits and 20+ free products, which covers a proof of concept before any spend.

The catalogue is Gemini 3.1 Pro, Gemini 3 Pro Image and Gemini 2.5 Pro. It runs from Pro tiers down to Flash and Flash-Lite, plus Live, Transcribe, Translate, Embedding, image, video and audio entries.

The largest context window is 1,048,576, ranked 11 of 72 among measured brands. That suits long documents and large codebases sent in a single request, rather than chunking text across calls.

The OpenAI-compatible API is yes, so a client written against that shape can be pointed at the Gemini endpoint. Your existing integration work carries over.

Input fees of $0.25 to $4 sit above the €0.07 site median, and output fees of $1.50 to $18 sit above the €0.23. Routing services can come in cheaper, as Salad shows at $0.06 on input.

Support runs on help centre and community, so questions go to the help centre or to community channels rather than to a named account contact. Plan escalation paths if your team relies on direct help.

Pick it for a single vendor's catalogue, Gemini 3.1 Pro, Gemini 3 Pro Image and Gemini 2.5 Pro, plus the OpenAI-compatible API. Pick Supascale AI for breadth, since it lists DeepSeek V4 Pro, GPT 5.6 Sol and GPT 5.6 Terra.

If you need OpenAI models instead, Azure AI Foundry carries gpt 6 astra, gpt 6 luna and gpt 6 sol, a different line this brand does not offer. Compare the two catalogues before switching.