GigaChat Integration Guide

GigaChat Integration Guide

Integration guide for GigaChat API (Sber Russian LLM) with CodeGraph.

Table of Contents

Overview

GigaChat is a Russian LLM provider from Sberbank. CodeGraph supports GigaChat as an alternative to OpenAI, Yandex AI Studio, or local models.

Tip: For larger context windows (up to 262K tokens), consider using Yandex AI Studio which offers Qwen3-235B and other models.

Quick Setup (3 Steps)

Step 1: Set Authorization Key

# Windows PowerShell
$env:GIGACHAT_AUTH_KEY = "your_authorization_key"

# Permanent (survives restart)
[System.Environment]::SetEnvironmentVariable('GIGACHAT_AUTH_KEY', 'your_key', 'User')

# Linux/Mac
export GIGACHAT_AUTH_KEY="your_authorization_key"

# Add to ~/.bashrc for permanent
echo 'export GIGACHAT_AUTH_KEY="your_key"' >> ~/.bashrc

Step 2: Configure config.yaml

llm:
  provider: gigachat

  gigachat:
    client_id: "019a7e2b-aeb3-78c4-ba3d-ddc1142b4ee6"
    scope: "GIGACHAT_API_PERS"
    model: "GigaChat-2-Pro"
    temperature: 0.7
    max_tokens: 2000
    timeout: 60

Step 3: Verify

python test_gigachat.py

Expected output:

GigaChat connection successful
Model: GigaChat-2-Pro
Status: Ready

Pre-configured Parameters

Parameter Value
Client ID 019a7e2b-aeb3-78c4-ba3d-ddc1142b4ee6
Scope GIGACHAT_API_PERS
Model GigaChat-2-Pro
Auth Key Set via environment variable

Available Models

Model Description Best For
GigaChat-2-Pro Most capable Complex analysis
GigaChat-2 Standard General use
GigaChat-Lite Fastest Simple queries

Automated Setup

PowerShell Script

# Run automated setup
.\setup_gigachat.ps1

The script will: 1. Prompt for Authorization Key 2. Create config.yaml from template 3. Install GigaChat SDK 4. Verify configuration

Using Template

# Copy template
cp config.gigachat.yaml.example config.yaml

# Edit with your settings
notepad config.yaml

Usage in Code

Basic Usage

import os
from langchain_gigachat import GigaChat

# Get auth key from environment
auth_key = os.getenv("GIGACHAT_AUTH_KEY")

# Initialize client
llm = GigaChat(
    credentials=auth_key,
    scope="GIGACHAT_API_PERS",
    model="GigaChat-2-Pro",
    verify_ssl_certs=False
)

# Make request
response = llm.invoke("What is PostgreSQL?")
print(response.content)

With CodeGraph

from src.llm.gigachat_provider import GigaChatProvider

# Initialize provider
provider = GigaChatProvider()

# Use with workflow
from src.workflow.langgraph_workflow_simple import run_workflow

result = run_workflow("Find transaction handling methods")
print(result['answer'])

Configuration Reference

Full config.yaml Example

llm:
  provider: gigachat

  gigachat:
    # Required
    client_id: "019a7e2b-aeb3-78c4-ba3d-ddc1142b4ee6"
    scope: "GIGACHAT_API_PERS"
    model: "GigaChat-2-Pro"

    # Optional
    temperature: 0.7          # Creativity (0.0-1.0)
    max_tokens: 2000          # Max response length
    timeout: 60               # Request timeout (seconds)
    verify_ssl_certs: false   # SSL verification
    profanity_check: false    # Content filter

    # Rate limiting
    max_retries: 3
    retry_delay: 1.0

Environment Variables

# Required
GIGACHAT_AUTH_KEY=your_authorization_key

# Optional
GIGACHAT_CLIENT_ID=019a7e2b-aeb3-78c4-ba3d-ddc1142b4ee6
GIGACHAT_SCOPE=GIGACHAT_API_PERS
GIGACHAT_MODEL=GigaChat-2-Pro

Troubleshooting

Authentication Failed (401)

Error: 401 Unauthorized

Solution:

# Check if key is set
echo $GIGACHAT_AUTH_KEY

# Verify key format (should be base64)
python -c "import base64; base64.b64decode('$GIGACHAT_AUTH_KEY')"

Connection Timeout

Error: Connection timed out

Solution:

# Increase timeout in config.yaml
gigachat:
  timeout: 120  # seconds

SSL Certificate Error

Error: SSL: CERTIFICATE_VERIFY_FAILED

Solution:

# Disable SSL verification (not recommended for production)
gigachat:
  verify_ssl_certs: false

Rate Limiting

Error: 429 Too Many Requests

Solution:

gigachat:
  max_retries: 5
  retry_delay: 2.0  # seconds

Best Practices

  1. Never commit auth keys - Use environment variables
  2. Use GigaChat-2-Pro - Best quality for code analysis
  3. Set appropriate timeout - 60s for normal, 120s for complex
  4. Enable retries - Handle transient failures

Security Considerations

  • Store auth keys in environment variables
  • Add .env to .gitignore
  • Use secure connection (HTTPS)
  • Consider SSL verification in production

Resources

Next Steps