SynapseX CLI

The official command-line interface for interacting with SynapseX services directly from your terminal.

Installation

pip install synapsex-cli

Or with pipx (recommended for CLI tools):

pipx install synapsex-cli

Configuration

Interactive Setup

synapsex configure

This prompts for:

API Key
License Key (optional)
Default model
Output format preference

Manual Configuration

# Set API key
synapsex configure --api-key sk-synapsex-your-key

# Set license key
synapsex configure --license-key SQXL-XXXX-XXXX-XXXX

# Check configuration
synapsex configure --show

Environment Variables

export SYNAPSEX_API_KEY="sk-synapsex-your-key"
export SYNAPSEX_LICENSE_KEY="SQXL-XXXX-XXXX-XXXX"
export SYNAPSEX_MODEL="synapsex-14b"

Commands Overview

Command	Description
`synapsex chat`	Chat with AI models
`synapsex train`	Manage training jobs
`synapsex dataset`	Manage datasets
`synapsex models`	List available models
`synapsex license`	License management
`synapsex feedback`	Submit feedback
`synapsex configure`	Configure credentials
`synapsex version`	Show version info

Chat Commands

Basic Chat

# Simple question
synapsex chat "What is machine learning?"

# Multi-line prompt
synapsex chat "Explain this code:
def fibonacci(n):
    if n <= 1:
        return n
    return fibonacci(n-1) + fibonacci(n-2)"

Chat Options

synapsex chat "Your prompt" [OPTIONS]

Option	Short	Description	Default
`--model`	`-m`	Model to use	`synapsex-v1`
`--temperature`	`-t`	Creativity (0.0-2.0)	`0.7`
`--max-tokens`		Max response length	`2048`
`--system`	`-s`	System prompt	None
`--rerank`	`-r`	Reranking mode	`none`
`--no-stream`		Disable streaming	Streaming on
`--json`	`-j`	JSON output	Rich format

Examples

# Use specific model
synapsex chat "Hello" --model synapsex-14b

# Creative writing
synapsex chat "Write a poem about AI" --temperature 1.2

# With system prompt
synapsex chat "Translate to Portuguese" \
  --system "You are a professional translator."

# Quantum reranking for complex queries
synapsex chat "Analyze market trends" --rerank quantum_gpu

# JSON output for scripting
synapsex chat "Hello" --json | jq '.content'

Reranking Modes

Mode	Description	Speed	Quality
`none`	No reranking	⚡⚡⚡	★★★
`cpu`	CPU reranking	⚡⚡	★★★★
`quantum_gpu`	Quantum GPU	⚡	★★★★★

Interactive Mode

# Start interactive chat session
synapsex chat --interactive

# With specific model
synapsex chat --interactive --model synapsex-14b

In interactive mode:

Type messages and press Enter
Use /clear to reset conversation
Use /model <name> to switch models
Use /exit or Ctrl+C to quit

Training Commands

Start Training

synapsex train start [OPTIONS]

Option	Short	Description	Required
`--type`	`-t`	Type: `sft`, `dpo`, `lora`	Yes
`--model`	`-m`	Base model	No
`--dataset`	`-d`	Dataset ID	No
`--gpus`	`-g`	Number of GPUs	No
`--epochs`	`-e`	Training epochs	No
`--batch-size`		Batch size	No
`--learning-rate`		Learning rate	No

Training Examples

# SFT training
synapsex train start --type sft --model synapsex-14b

# LoRA with 4 GPUs
synapsex train start --type lora --gpus 4 --epochs 5

# DPO with custom dataset
synapsex train start --type dpo --dataset my-preferences

Monitor Training

# Job status
synapsex train status <job-id>

# Watch progress (real-time)
synapsex train status <job-id> --watch

# List all jobs
synapsex train list

# Filter by status
synapsex train list --status running
synapsex train list --status completed
synapsex train list --status failed

Training Logs

# View logs
synapsex train logs <job-id>

# Follow logs in real-time
synapsex train logs <job-id> --follow

# Last 100 lines
synapsex train logs <job-id> --tail 100

Cancel Training

synapsex train cancel <job-id>

Dataset Commands

Upload Dataset

# Upload JSONL file
synapsex dataset upload ./data.jsonl --name "my-dataset"

# With description
synapsex dataset upload ./data.jsonl \
  --name "support-v2" \
  --description "Customer support conversations"

List Datasets

synapsex dataset list

# Detailed view
synapsex dataset list --verbose

Dataset Info

synapsex dataset info <dataset-id>

Delete Dataset

synapsex dataset delete <dataset-id>

# Force delete (no confirmation)
synapsex dataset delete <dataset-id> --force

Download Dataset

synapsex dataset download <dataset-id> --output ./download.jsonl

Models Commands

List Models

synapsex models list

# Show all details
synapsex models list --verbose

Model Info

synapsex models info synapsex-14b

Output:

Model: synapsex-14b
Parameters: 14B
Context Length: 32768 tokens
Capabilities: chat, code, analysis
Pricing: $0.0005/1K tokens

License Commands

Check License

synapsex license info

Output:

Plan: Professional
Status: Active
Expires: 2026-12-31
Features:
  ✓ Quantum Reranking
  ✓ Custom Training
  ✓ Priority Support
Quota:
  API Calls: 45,000/100,000 (45%)
  Training Hours: 12/50 (24%)

Validate License

synapsex license validate

Activate License

synapsex license activate SQXL-XXXX-XXXX-XXXX

Feedback Commands

Submit Feedback

# Quick rating
synapsex feedback submit --request-id req-123 --rating 5

# With comment
synapsex feedback submit \
  --request-id req-123 \
  --rating 4 \
  --comment "Good but could be more concise"

List Feedback

synapsex feedback list --limit 10

Global Options

These options work with all commands:

Option	Description
`--help`	Show help for command
`--version`	Show CLI version
`--verbose`	Verbose output
`--quiet`	Minimal output
`--json`	JSON output format
`--no-color`	Disable colored output

Output Formats

Default (Rich)

synapsex chat "Hello"

╭─ SynapseX Response ─────────────────────────────╮
│ Hello! How can I assist you today?             │
│                                                 │
│ Model: synapsex-v1  Tokens: 23                 │
╰─────────────────────────────────────────────────╯

JSON

synapsex chat "Hello" --json

{
  "content": "Hello! How can I assist you today?",
  "model": "synapsex-v1",
  "usage": {
    "prompt_tokens": 8,
    "completion_tokens": 15,
    "total_tokens": 23
  }
}

Plain

synapsex chat "Hello" --plain

Hello! How can I assist you today?

Scripting Examples

Batch Processing

#!/bin/bash
# Process multiple files

for file in ./prompts/*.txt; do
    prompt=$(cat "$file")
    output="${file%.txt}.response.txt"
    synapsex chat "$prompt" --plain > "$output"
    echo "Processed: $file"
done

Pipeline Integration

# Use with pipes
echo "Summarize this article" | synapsex chat --stdin

# Chain commands
cat document.txt | synapsex chat "Translate to Spanish:" --stdin --plain > spanish.txt

JSON Processing

# Extract specific fields
synapsex chat "List 5 programming languages" --json | \
  jq -r '.content' | \
  grep -E "^[0-9]+\."

Monitoring Script

#!/bin/bash
# Watch training job

JOB_ID="$1"
while true; do
    status=$(synapsex train status "$JOB_ID" --json | jq -r '.status')
    progress=$(synapsex train status "$JOB_ID" --json | jq -r '.progress')
    
    echo "[$status] Progress: $progress%"
    
    if [[ "$status" == "completed" ]] || [[ "$status" == "failed" ]]; then
        break
    fi
    
    sleep 60
done

Troubleshooting

Common Issues

Authentication Failed

# Check configuration
synapsex configure --show

# Reset and reconfigure
synapsex configure --reset
synapsex configure

Connection Timeout

# Increase timeout
synapsex chat "Hello" --timeout 60

Rate Limited

# Check current usage
synapsex license info

# Wait and retry (automatic with --retry)
synapsex chat "Hello" --retry 3

Debug Mode

# Enable debug output
synapsex chat "Hello" --debug

# Very verbose
synapsex chat "Hello" -vvv

Log Location

# View CLI logs
cat ~/.synapsex/cli.log

# Tail logs
tail -f ~/.synapsex/cli.log

Configuration File

Location: ~/.synapsex/config.yaml

api_key: sk-synapsex-your-key
license_key: SQXL-XXXX-XXXX-XXXX
defaults:
  model: synapsex-14b
  temperature: 0.7
  max_tokens: 2048
  rerank_mode: none
output:
  format: rich  # rich, json, plain
  color: true
  stream: true
api:
  base_url: https://api.synapsex.ai
  timeout: 30
  retries: 3

Installation​

Configuration​

Interactive Setup​

Manual Configuration​

Environment Variables​

Commands Overview​

Chat Commands​

Basic Chat​

Chat Options​

Examples​

Reranking Modes​

Interactive Mode​

Training Commands​

Start Training​

Training Examples​

Monitor Training​

Training Logs​

Cancel Training​

Dataset Commands​

Upload Dataset​

List Datasets​

Dataset Info​

Delete Dataset​

Download Dataset​

Models Commands​

List Models​

Model Info​

License Commands​

Check License​

Validate License​

Activate License​

Feedback Commands​

Submit Feedback​

List Feedback​

Global Options​

Output Formats​

Default (Rich)​

JSON​

Plain​

Scripting Examples​

Batch Processing​

Pipeline Integration​

JSON Processing​

Monitoring Script​

Troubleshooting​

Common Issues​

Debug Mode​

Log Location​

Configuration File​

See Also​