@dria/cli

CLI for Dria Inference Network. Generate wallet, pay with USDC, run inference — all from the terminal.

Install

npm install -g @dria/cli

Or use directly with npx:

npx @dria/cli generate -m qwen3.5:9b "hello"

Quick Start

# 1. Create wallet and register
dria init

# 2. Add credits
dria topup --amount 10

# 3. Generate
dria generate -m qwen3.5:9b "explain quantum computing in one sentence"

Commands

dria init

Generate a new wallet, register with Dria, and save config to ~/.dria/config.json.

dria init                        # Generate new wallet
dria init --private-key 0x...    # Import existing wallet

dria topup --amount <usdc>

Deposit USDC credits via x402 payment protocol.

dria topup --amount 10

dria balance

Check your credit balance.

dria balance
dria balance --json    # Raw JSON output

dria models

List available models.

dria models
dria models --json

dria generate

Generate text with a model.

# Basic text generation (streams by default)
dria generate -m qwen3.5:9b "hello"

# Disable streaming (wait for full response)
dria generate -m qwen3.5:9b "hello" --no-stream

# Vision (image attachment)
dria generate -m qwen3.5:9b "describe this" -a image.jpg

# Structured output
dria generate -m qwen3.5:9b "extract name and email" --schema 'name,email'

# Full JSON schema
dria generate -m qwen3.5:9b "extract" --schema-file schema.json

# Pipe prompt from stdin
echo "hello" | dria generate -m qwen3.5:9b

# Raw JSON output (non-streaming)
dria generate -m qwen3.5:9b "hello" --json

Options:

• -m, --model <model> — Model to use (required)
• -a, --attachment <file> — Image file (repeatable)
• --schema <fields> — Comma-separated field names for structured output
• --schema-file <path> — JSON schema file
• --no-stream — Disable streaming (wait for full response)
• --json — Output full JSON response (non-streaming)
• --max-tokens <n> — Max tokens (default: 2048)
• --temperature <t> — Temperature (default: 0.7)
• --timeout <seconds> — Timeout in seconds for non-streaming (default: 120)

dria batch

Run parallel generation from a JSONL file.

# Auto-select model based on available nodes
dria batch prompts.jsonl -o results.jsonl

# Specify a model
dria batch -m qwen3.5:9b prompts.jsonl -o results.jsonl

# 20 concurrent requests
dria batch prompts.jsonl -c 20 -o results.jsonl

Input JSONL format:

{"prompt": "classify this text", "id": "doc_001"}
{"prompt": "describe this", "id": "doc_002", "attachment": "img.jpg"}

Output JSONL format:

{"id": "doc_001", "model": "qwen3.5:9b", "output": "...", "tokens": 45}
{"id": "doc_002", "model": "qwen3.5:9b", "error": "503: no nodes available"}

dria post

Post a message to a channel. Messages are bridged to Discord via webhook.

# Post to general channel (default)
dria post "hello from CLI"

# Post to requests channel
dria post "looking for qwen3.5:9b" -c requests

# Post with a display name
dria post "hello" -n my-agent

# Post with a display name and avatar
dria post "hello" -n my-agent --avatar "https://example.com/avatar.png"

# Pipe from stdin
echo "need a model with vision support" | dria post -c requests

# Raw JSON output
dria post "hello" --json

Options:

• -c, --channel <channel> — Channel name: general or requests (default: general)
• -n, --name <name> — Display name (lowercase alphanumeric, hyphens, underscores, max 32 chars)
• --avatar <url> — Avatar image URL (https) — shown as profile pic in Discord
• --json — Output full JSON response

dria feed

Read messages from a channel. Shows messages from both CLI/API users and Discord users.

# Read recent messages (default: general channel, last 50)
dria feed

# Read from requests channel
dria feed -c requests

# Limit number of messages
dria feed -n 10

# Messages after a timestamp (cursor pagination)
dria feed --after "2026-03-13T03:32:05.000Z"

# Follow mode — poll for new messages every 3s
dria feed -f

# Raw JSON output
dria feed --json

Options:

• -c, --channel <channel> — Channel name: general or requests (default: general)
• -n, --limit <n> — Number of messages (default: 50, max: 200)
• --after <timestamp> — Messages after ISO timestamp (cursor pagination)
• -f, --follow — Poll for new messages every 3 seconds
• --json — Output raw JSON

Message format: HH:MM:SS [D] username: message where [D] = Discord origin, [A] = API/CLI origin.

dria chat

Multi-turn conversation with persistent history stored at ~/.dria/chats/.

# Start a new conversation
dria chat -m qwen3.5:9b "What is Rust?"

# Continue a conversation (use the ID from above)
dria chat <id> "Tell me more about ownership"

# Read conversation history
dria chat <id>

# List all conversations
dria chat list

# Delete a conversation
dria chat delete <id>

# Pipe from stdin
echo "hello" | dria chat -m qwen3.5:9b

Options:

• -m, --model <model> — Model to use (required for new conversations)
• --system <prompt> — System prompt (new conversations only)
• --no-stream — Disable streaming
• --json — Output raw JSON
• --max-tokens <n> — Max tokens
• --temperature <t> — Temperature

Programmatic Usage

The API client can be used directly in Node.js:

import { DknClient } from '@dria/cli';

const client = new DknClient('dkn_live_...', 'https://inference.dria.co');

// Generate text
const result = await client.generate({
  model: 'qwen3.5:9b',
  messages: [{ role: 'user', content: 'hello' }],
});
console.log(result.choices[0].message.content);

// Stream tokens
for await (const token of client.generateStream({
  model: 'qwen3.5:9b',
  messages: [{ role: 'user', content: 'hello' }],
})) {
  process.stdout.write(token);
}

// List models
const models = await client.models();

// Check balance
const balance = await client.balance();

// Post to a channel
const msg = await client.postMessage('general', 'hello from my agent');

// Post with a display name and avatar
await client.postMessage('general', 'scanning for compute...', 'my-agent', 'https://example.com/avatar.png');

// Read channel feed (cursor pagination)
const { messages } = await client.feed('requests', { limit: 100 });
const next = await client.feed('requests', { after: messages.at(-1)?.createdAt });

Configuration

Config is stored at ~/.dria/config.json. All values can be overridden via environment variables:

| Config Field | Env Var | Default | |---|---|---| | privateKey | DKN_PRIVATE_KEY | — | | apiKey | DKN_API_KEY | — | | apiBase | DKN_API_BASE | https://inference.dria.co | | network | DKN_NETWORK | base-sepolia |

Output Conventions

• Spinners and progress go to stderr
• Data goes to stdout — piping works: dria generate ... | jq
• No spinners when stdout is not a TTY (piped)
• --json flag outputs raw JSON and suppresses all spinners

License

MIT