Developer documentation

One gateway. Every model.

Neosantara is a unified AI gateway: call 100+ models — GPT, Claude, Gemini, and more — through one API key and OpenAI-compatible endpoints. Building with an AI coding agent instead? Subscribe to the Coding Plan for flat-fee, prompt-based access in Claude Code, Cline, or Jelma.

Start quickstart Browse models

Start here

Create a key and make your first request.

Compare live models, capabilities, and pricing.

Subscribe for prompt-based access in your coding agent.

Browse endpoint schemas and try requests live.

Choose a path

Five ways to build with Neosantara.

Chat & multimodal API

Send chat, embeddings, image, audio, and file requests with SDKs you already know.

Coding Plan

Flat-fee prompt quota for Claude Code, Cline, and Jelma — no per-token billing.

Framework integrations

Connect through native and provider adapters when you want agents, tools, orchestration, or one interface across providers.

Usage and limits

Understand request limits, model costs, token accounting, and how usage behaves across tiers.

Endpoint reference

Jump into generated request schemas, parameters, response bodies, and playground examples.

Integration spotlight

Two native integrations, ready to use.

Agents

Agno native

Use agno.models.neosantara.Neosantara for multimodal agents, streaming, tools, and model switching.

Provider layer

Any-LLM native

Use provider id neosantara with NEOSANTARA_API_KEY for chat completions, Responses, images, and batch workflows.

First request

Start from the API, then move into the framework you prefer.

OpenAI-compatible request

Call a model with your API key.

curl https://api.neosantara.xyz/v1/responses \
  -H "Authorization: Bearer $NEOSANTARA_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "gemini-3.5-flash",
    "input": "Halo Neosantara"
  }'

import os
from openai import OpenAI

client = OpenAI(
    api_key=os.environ["NEOSANTARA_API_KEY"],
    base_url="https://api.neosantara.xyz/v1",
)

response = client.responses.create(
    model="gemini-3.5-flash",
    input="Halo Neosantara",
)

print(response.output_text)

Getting Started

SDKs & APIs

Core Concepts

Jelma

Capabilities