Proxy Server Settings

Configure the built-in API proxy server for OpenAI/Anthropic/Gemini compatibility

Overview

Antigravity Manager includes a built-in proxy server that translates OpenAI, Anthropic, and Gemini API requests to use your Google AI accounts.

Basic Configuration

Server Settings

enabled · boolean · default: "false"

Enable the proxy service

port · number · default: "8045"

Listening port for the proxy server

Range: 8000-65535

Location: config.rs:478

auto_start · boolean · default: "false"

Automatically start proxy server when application launches

Location: config.rs:487

request_timeout · number · default: "120"

API request timeout in seconds

Default: 120 seconds
Range: 30-7200 seconds

Location: config.rs:494-495

Network Access

allow_lan_access · boolean · default: "false"

Allow LAN access to the proxy server

• false: Bind to 127.0.0.1 (local only, privacy-first)
• true: Bind to 0.0.0.0 (allow LAN access)

Location: config.rs:463-467

When enabling LAN access, ensure you configure authentication to prevent unauthorized access.

Authentication Modes

auth_mode · enum · default: "auto"

API authentication policy

Options:

• off: No authentication required
• strict: Authentication required for all routes
• all_except_health: Authentication required except /healthz
• auto: Recommended defaults (LAN=all_except_health, local=off)

Location: config.rs:469-476

api_key · string · required

API key for client authentication

Format: Must start with sk- followed by UUID

Default: Auto-generated on first run
Example: sk-a1b2c3d4e5f6...

Location: config.rs:481

admin_password · string

Web UI management console password

If not set, defaults to api_key value.

Location: config.rs:484

Upstream Proxy Configuration

Route requests through an upstream proxy server:

upstream_proxy.enabled · boolean · default: "false"

Enable upstream proxy

Location: config.rs:564

upstream_proxy.url · string

Upstream proxy URL

Supported protocols: http://, https://, socks5://

Example: http://127.0.0.1:7890

Location: config.rs:566

If the proxy URL doesn’t include a protocol, http:// is automatically prepended. See normalize_proxy_url() in config.rs:10-21

Logging Configuration

enable_logging · boolean · default: "true"

Enable request logging (required for token statistics)

Location: config.rs:499

debug_logging.enabled · boolean · default: "false"

Enable debug logging (saves full request/response chain)

Location: config.rs:502

debug_logging.output_dir · string

Custom directory for debug logs

If not set, uses default application data directory.

Location: config.rs:503

Advanced Settings

User-Agent Override

user_agent_override · string

Override User-Agent header sent to upstream APIs

Example: antigravity/1.15.8 darwin/arm64

Location: config.rs:515

saved_user_agent · string

Persisted User-Agent value (even when override is disabled)

Location: config.rs:536-537

Configuration Example

{
  "proxy": {
    "enabled": true,
    "port": 8045,
    "auto_start": true,
    "allow_lan_access": false,
    "auth_mode": "auto",
    "api_key": "sk-a1b2c3d4e5f6...",
    "admin_password": "my-secure-password",
    "request_timeout": 120,
    "enable_logging": true,
    "upstream_proxy": {
      "enabled": false,
      "url": ""
    },
    "debug_logging": {
      "enabled": false,
      "output_dir": null
    },
    "user_agent_override": null
  }
}

Usage Examples

Python (OpenAI SDK)

from openai import OpenAI

client = OpenAI(
    base_url="http://127.0.0.1:8045/v1",
    api_key="sk-your-antigravity-key"
)

response = client.chat.completions.create(
    model="gemini-3-flash",
    messages=[{"role": "user", "content": "Hello!"}]
)

print(response.choices[0].message.content)

Python (Anthropic SDK)

from anthropic import Anthropic

client = Anthropic(
    base_url="http://127.0.0.1:8045",
    api_key="sk-your-antigravity-key"
)

response = client.messages.create(
    model="claude-sonnet-4-6",
    max_tokens=1024,
    messages=[{"role": "user", "content": "Hello!"}]
)

print(response.content[0].text)

Python (Native Gemini SDK)

import google.generativeai as genai

genai.configure(
    api_key="sk-your-antigravity-key",
    transport='rest',
    client_options={'api_endpoint': 'http://127.0.0.1:8045'}
)

model = genai.GenerativeModel('gemini-3-flash')
response = model.generate_content("Hello!")
print(response.text)