XavaInferencesign indocs
docs

Agent plugins

One plugin for every major coding agent: our MCP server, five skills, slash commands and a model-picker subagent. Same key, same balance.

The plugin lives at github.com/xavalabs/xinf-plugin. One canonical bundle, with a manifest per client. It adds our MCP server and teaches your agent when and how to use it. The marketing overview is on /agents.

  1. Run the install for your client below.
  2. Set XINF_API_KEY to a key from your dashboard (/xinf:setup walks you through it). Without a key, the catalog, pricing and x402 media tools still work.
  3. Restart the agent, then try /xinf:balance or ask it for an image.

The plugin defaults to our production origin. To use this deployment instead, set export XINF_BASE_URL=https://zinf.dev before starting the agent.

What is included

partnamewhat it does
skillusing-xinfpick a model, OpenAI-compatible base URL, keys, balance, errors
skillxinf-mediagenerate images, video and audio through the tools or /v1, and return the URLs
skillxinf-x402pay per media call in USDC: 402, sign, retry; the X-Split header; budget caps
skillxinf-models-and-pricingstandard price vs your price, the liquidity book, cost-effective choices
skillxinf-buybackshow an account's split works, and $XINF
commands/xinf:models /xinf:balance /xinf:usage /xinf:generate-image /xinf:setupslash commands (Claude Code, Cursor, Kimi)
subagentmodel-pickerrecommends a model for a task by price and capability; read-only

x402 budget caps

When the agent pays per call, the xinf-x402 skill makes it state each price and never pay above your cap without asking. Set XINF_X402_MAX_PER_CALL_USD and XINF_X402_SESSION_BUDGET_USD, and use a dedicated wallet with a small USDC float.

Claude Code

Installs the plugin: the MCP server, five skills, /xinf commands and a model-picker subagent.

terminal
claude plugin marketplace add xavalabs/xinf-plugin && claude plugin install xinf@xinf
or only the MCP server
claude mcp add --transport http xinf https://zinf.dev/mcp \
  --header "Authorization: Bearer $XINF_API_KEY"

Codex

Installs the plugin in Codex; the MCP server reads your key from an environment variable.

terminal
codex plugin marketplace add xavalabs/xinf-plugin && codex plugin add xinf@xinf
or only the MCP server
codex mcp add xinf --url https://zinf.dev/mcp --bearer-token-env-var XINF_API_KEY
~/.codex/config.toml (optional: run Codex itself on our models)
[model_providers.xinf]
name = "Xava Inference"
base_url = "https://zinf.dev/v1"
env_key = "XINF_API_KEY"
wire_api = "responses"

[profiles.xinf]
model_provider = "xinf"
model = "openai/gpt-5.4"

# then: codex --profile xinf

Codex speaks the Responses API, which we serve for OpenAI models.

Cursor

Adds the plugin to Cursor's agent: skills, commands and the MCP server.

in Cursor chat
/add-plugin https://github.com/xavalabs/xinf-plugin
or only the MCP server (~/.cursor/mcp.json)
{
  "mcpServers": {
    "xinf": {
      "url": "https://zinf.dev/mcp",
      "headers": { "Authorization": "Bearer ${env:XINF_API_KEY}" }
    }
  }
}
or the one-click MCP install link
cursor://anysphere.cursor-deeplink/mcp/install?name=xinf&config=eyJ1cmwiOiJodHRwczovL3ppbmYuZGV2L21jcCIsImhlYWRlcnMiOnsiQXV0aG9yaXphdGlvbiI6IkJlYXJlciAke2VudjpYSU5GX0FQSV9LRVl9In19
chat models (Settings > Models)
Override OpenAI Base URL:  https://zinf.dev/v1
OpenAI API Key:            xk_live_...
Add model:                 anthropic/claude-sonnet-5

Gemini CLI

Installs the extension; it sends XINF_API_KEY from your environment as the bearer.

terminal
gemini extensions install https://github.com/xavalabs/xinf-plugin --consent
or only the MCP server
gemini mcp add --transport http \
  --header "Authorization: Bearer $XINF_API_KEY" xinf https://zinf.dev/mcp

OpenCode

Installs the plugin globally; with a key set it also registers us as a model provider.

terminal
curl -fsSL https://raw.githubusercontent.com/xavalabs/xinf-plugin/main/scripts/install-opencode.sh | sh
or by hand (opencode.json)
{
  "mcp": {
    "xinf": {
      "type": "remote",
      "url": "https://zinf.dev/mcp",
      "enabled": true,
      "headers": { "Authorization": "Bearer {env:XINF_API_KEY}" }
    }
  },
  "provider": {
    "xinf": {
      "npm": "@ai-sdk/openai-compatible",
      "name": "Xava Inference",
      "options": { "baseURL": "https://zinf.dev/v1", "apiKey": "{env:XINF_API_KEY}" },
      "models": { "anthropic/claude-sonnet-5": {}, "openai/gpt-5.4": {} }
    }
  }
}

Kimi

Installs the plugin in Kimi Code; it applies in a new session.

in Kimi chat
/plugins install https://github.com/xavalabs/xinf-plugin

Pi

Installs the Pi package: the skills plus the MCP server, added to your Pi MCP config.

terminal
pi install git:github.com/xavalabs/xinf-plugin@v0.1.0

Cline

Paste into MCP Servers > Configure, with your key (Cline has no environment variables here).

cline_mcp_settings.json
{
  "mcpServers": {
    "xinf": {
      "type": "streamableHttp",
      "url": "https://zinf.dev/mcp",
      "headers": { "Authorization": "Bearer xk_live_..." }
    }
  }
}
chat models (API Provider: OpenAI Compatible)
Base URL:  https://zinf.dev/v1
API Key:   xk_live_...
Model ID:  anthropic/claude-sonnet-5

Windsurf

Merge into Cascade's MCP config (~/.codeium/windsurf/mcp_config.json; newer builds ~/.config/devin/mcp_config.json).

mcp_config.json
{
  "mcpServers": {
    "xinf": {
      "serverUrl": "https://zinf.dev/mcp",
      "headers": { "Authorization": "Bearer ${env:XINF_API_KEY}" }
    }
  }
}

Any OpenAI SDK

Two lines: every OpenAI-compatible SDK and framework picks these up.

.env
OPENAI_BASE_URL=https://zinf.dev/v1
OPENAI_API_KEY=xk_live_...