Google Vertex AI via VerticalAPI

Supported models

Google Vertex AI models routed by VerticalAPI

Pass the model ID below as model in any OpenAI-compatible request. New Google Vertex AI models are typically supported within 24h of release.

Model ID	Name	Context	Pricing (provider)
`gemini-2.5-pro`	Gemini 2.5 Pro (Vertex)	2M	Vertex pricing
`claude-sonnet-4-5@vertex`	Claude Sonnet 4.5 (Vertex)	200K	Anthropic-on-Vertex pricing
`llama-3.3-70b@vertex`	Llama 3.3 70B (Vertex)	128K	Vertex Llama pricing

Pricing reflects Google Vertex AI's rates — you pay Google Vertex AI directly. VerticalAPI adds zero markup on tokens.

Quickstart

5-line Google Vertex AI call via VerticalAPI

Drop-in replacement for the OpenAI SDK. Works with the OpenAI Python client, Node, Go, curl — anything that speaks HTTP.

                vertex-ai_quickstart.py
                Python
            
from openai import OpenAI

client = OpenAI(
    base_url="https://api.verticalapi.com/v1",
    api_key="vapi_...",
    default_headers={"X-Provider-Key": "GCP service-ac..."}
)

response = client.chat.completions.create(
    model="gemini-2.5-pro",  # Google Vertex AI
    messages=[{"role": "user", "content": "Hello"}]
)
print(response.choices[0].message.content)

Google Vertex AI and VerticalAPI

What VerticalAPI does and does not do with Google Vertex AI

Not routed today

VerticalAPI does not proxy Google Vertex AI. This page documents Google Vertex AI's own models and pricing so you can compare it; call Google Vertex AI directly to use it.

What it does route

11 providers behind one OpenAI-compatible endpoint: Anthropic, Cerebras, DeepInfra, Fireworks, Google AI Studio, Groq, Mistral, OpenAI, OpenRouter, Together, xAI. Switchable per request, same SDK.

No token markup on those

Bring your own key and pay the provider directly. VerticalAPI's revenue is the gateway subscription, not a share of your tokens.

Prompts are never stored

Each request is recorded as metadata only — vertical, model, token counts, latency, status. There is no column for prompt or completion text.

Best for

Where Google Vertex AI shines

GCP-native compliance data residency consolidated GCP billing Anthropic-on-GCP

Frequently asked questions

What is Vertex AI and what models do they offer?

Vertex AI is Google Cloud's managed AI/ML platform. The 2026 model garden includes Gemini 2.5 Pro and Flash, Anthropic Claude 4.5 (via Anthropic-on-Vertex), Meta Llama 3.3, Mistral Large, Imagen 3 for images, Veo 2 for video, Chirp 3 for speech, plus custom-tuned Gemini and open-source models on Model Garden. Vertex adds MLOps tooling, agent builders, RAG (Vertex Search) and evaluation.

How much does Vertex AI cost in 2026?

Gemini 2.5 Pro is $1.25 per 1M input tokens (up to 200K) and $10 per 1M output. Gemini 2.5 Flash is around $0.30/$2.50. Claude Sonnet 4.5 on Vertex matches Anthropic's $3/$15. Llama 3.3 70B on Vertex is roughly $0.72/$0.72. Plus Google Cloud egress and Vertex feature fees. You pay Vertex AI directly, on Vertex AI's own account.

Can I use Vertex AI through VerticalAPI?

Create a Google Cloud service account with aiplatform.user role, download the JSON key, paste it into VerticalAPI (with project ID and region), then point the OpenAI SDK at https://api.verticalapi.com/v1. VerticalAPI handles GCP auth (signed tokens), translates OpenAI chat completions into Vertex generateContent, and preserves multimodal inputs. Billing remains on your Google Cloud invoice.

What is Vertex AI best for compared to alternatives?

Vertex AI wins for Google Cloud enterprises and for 2M-context multimodal workloads: Gemini 2.5 Pro's video/audio/PDF understanding is unique, and Vertex adds IAM, VPC-SC, CMEK and regional residency. Compared to AI Studio (consumer Gemini API) it adds enterprise controls. Compared to AWS Bedrock or Azure OpenAI it leads on multimodal context but lags on third-party model breadth (Claude is recent, OpenAI absent).

Where is Vertex AI hosted / data privacy?

Vertex AI runs in 20+ Google Cloud regions including us-central1, europe-west1/4, asia-northeast1 and sovereign EU regions. Data is not used to train models. VPC Service Controls, customer-managed encryption keys, Access Transparency and EU Data Residency are available. Your contract and data terms stay between you and Vertex AI.

Limitations and trade-offs

Authentication is heavier than API keys — requires Google Cloud project, IAM, service accounts.
Model availability per region varies — Gemini Pro is global but Claude on Vertex is US/EU only.
Vertex egress and feature fees stack on top of token pricing — total cost can exceed list.
Quotas (QPM, TPM) are per region and per project and often need explicit increase requests.
Some open-weight models on Model Garden require manual deployment and pay-per-hour endpoint hosting.

Where Vertex AI is heading

Gemini 3 launching on Vertex with deeper agentic capabilities and tool use.
Wider third-party model availability (rumored Llama 4, OpenAI partnerships via partner gateways).
Expanded Agent Builder and Vertex AI Search for grounded RAG.
More sovereign EU and APAC region launches for regulated industries.

All supported LLM providers

Same endpoint, same SDK — just change the model and the BYOK header.

Ship on Google Vertex AI in 60 seconds

Free tier — bring your own Google Vertex AI key, zero markup, OpenAI-compatible endpoint.

Get your VerticalAPI key →