OpenCode integration guide

Developer toolCLI toolUpdated: 2026-04-17

Introduction

OpenCode is an open-source AI coding assistant that runs in your terminal, helping you write, debug and refactor code conversationally. Point it at the SeaWhale AI API and you can use Claude, GPT-5, Qwen and other powerful models.

Key capabilities

💻 Code generation — generate code from a natural language description
🐛 Debugging — locate and fix bugs quickly
♻️ Refactoring — improve code structure and performance
📝 Documentation — add comments and docs to your code
🔍 Code explanation — make sense of complex logic
🚀 Many languages — Python, JavaScript, Java, Go and more

Why SeaWhale AI?

Advantage	Description
🌐 Direct access	Reachable directly, stable connectivity
💰 Flexible billing	Pay as you go, no subscription required
⚡ High performance	Low latency, fast responses
🔒 Data security	Code is not stored, protecting your privacy
🤖 Many models	Claude, GPT-5, Qwen and more
🆓 New user credit	New accounts receive free credit

Supported models

Through the OpenAI-compatible endpoint, SeaWhale AI supports:

Family	Recommended model	Strengths	Best for
Claude Sonnet	`claude-sonnet-4-6`	Strong at code, accurate reasoning, 200K context	Complex projects, refactoring
Claude Opus	`claude-opus-4-7`	Highest capability, deep reasoning	Architecture, complex algorithms
GPT-5	`gpt-5.4`	Strong multimodal, well balanced	Everyday work, varied tasks
DeepSeek	`deepseek-v3`	Flagship open model, excellent at code	Coding tasks
Qwen	`qwen-max`	Strong in Chinese, fast responses	Chinese-language projects

Choosing a model

Everyday coding: claude-sonnet-4-6 (strong at code, good value)
Complex tasks: claude-opus-4-7 (highest capability)
Best value: deepseek-v3 (low cost, strong at code)

Before you begin

1. Get a SeaWhale AI API key

Open the SeaWhale AI console
Sign up and log in
Generate an API key on the API management page
Make sure your account has enough balance or free credit

New user credit

New SeaWhale AI accounts receive free starter credit, usable across all model inference services.

2. System requirements

Item	Requirement
Operating system	macOS 10.15+, Windows 10+, Linux
Node.js	v18.0 or later
npm	v7.0+
Terminal	A modern terminal with color support

Checking your Node.js version

bash

node -v

Version v18.x.x or later meets the requirement. If Node.js is missing, download it from the Node.js website.

Installing OpenCode

Global install

bash

npm install -g opencode-ai

Verify the installation

bash

opencode -v

A version number means the install succeeded.

Installation notes

If installation fails, check that Node.js is version 18 or later
To speed up npm, you can switch registries: npm config set registry https://registry.npmmirror.com
On macOS/Linux, permission errors can be resolved with sudo npm install -g opencode-ai

Configuring SeaWhale AI

Configuration file (recommended)

Create or edit the OpenCode configuration file:

macOS / Linux: ~/.config/opencode/opencode.json
Windows: C:\Users\YourName\.config\opencode\opencode.json

About the Base URL

baseURL must end with /v1, otherwise you get a 404 Not Found error.

Full configurationMinimal configuration

json

{
  "$schema": "https://opencode.ai/config.json",
  "provider": {
    "seawhale": {
      "npm": "@ai-sdk/openai-compatible",
      "name": "SeaWhale AI",
      "options": {
        "baseURL": "https://api.atalk-ai.com/v1",
        "apiKey": "sk-xxxxxxxxxxxxxxxx"
      },
      "models": {
        "claude-sonnet-4-6": {
          "name": "Claude Sonnet 4.6",
          "modalities": {
            "input": ["text", "image"],
            "output": ["text"]
          },
          "limit": {
            "context": 200000,
            "output": 64000
          }
        },
        "claude-opus-4-7": {
          "name": "Claude Opus 4.7",
          "modalities": {
            "input": ["text", "image"],
            "output": ["text"]
          },
          "limit": {
            "context": 200000,
            "output": 32000
          }
        },
        "deepseek-v3": {
          "name": "DeepSeek V3",
          "modalities": {
            "input": ["text"],
            "output": ["text"]
          },
          "limit": {
            "context": 128000,
            "output": 8192
          }
        },
        "gpt-5.4": {
          "name": "GPT-5.4",
          "modalities": {
            "input": ["text", "image"],
            "output": ["text"]
          },
          "limit": {
            "context": 128000,
            "output": 16384
          }
        },
        "qwen-max": {
          "name": "Qwen Max",
          "modalities": {
            "input": ["text"],
            "output": ["text"]
          },
          "limit": {
            "context": 32768,
            "output": 8192
          }
        }
      }
    }
  }
}

json

{
  "$schema": "https://opencode.ai/config.json",
  "provider": {
    "seawhale": {
      "npm": "@ai-sdk/openai-compatible",
      "name": "SeaWhale AI",
      "options": {
        "baseURL": "https://api.atalk-ai.com/v1",
        "apiKey": "sk-xxxxxxxxxxxxxxxx"
      },
      "models": {
        "claude-sonnet-4-6": {
          "name": "Claude Sonnet 4.6",
          "modalities": {
            "input": ["text", "image"],
            "output": ["text"]
          },
          "limit": {
            "context": 200000,
            "output": 64000
          }
        }
      }
    }
  }
}

Notes

Replace sk-xxxxxxxxxxxxxxxx with your real SeaWhale AI API key
Quit and restart OpenCode after editing the config for changes to take effect
Keep your API key safe and never commit it to a repository

Getting started

Starting OpenCode

bash

cd your-project    # Change into your project
opencode           # Start OpenCode

OpenCode chat interface

TIP

OpenCode uses the current directory as project context, so start it in the right place. If you do not have a project yet, run mkdir my-project && cd my-project first.

Common commands

Command	What it does
`/connect`	Connect or switch providers
`/models`	Browse and switch models
`/clear`	Clear conversation history
`/help`	Show help

Switching models

Type /models, search for SeaWhale AI or a model name (such as claude or deepseek), and select the model you want.

OpenCode model selection

Worked examples

Example 1: code generation

👤 User:
Write a Python function that parses a JSON file and returns the values of a given field

🤖 OpenCode:
def extract_field(filepath, field):
    import json
    with open(filepath, 'r', encoding='utf-8') as f:
        data = json.load(f)
    if isinstance(data, list):
        return [item.get(field) for item in data if isinstance(item, dict)]
    elif isinstance(data, dict):
        return data.get(field)
    return None

Example 2: debugging

👤 User:
This function raises KeyError — please fix it:
def get_user_name(user):
    return user['name']

🤖 OpenCode:
Use .get() for safe access with a default value:
def get_user_name(user):
    return user.get('name', 'Unknown user')

Saving tokens

1. Start in the specific project directory

Best practices

Start OpenCode inside the specific project directory, not a parent directory
Use .gitignore to exclude node_modules/, dist/ and similar
Delete or move large binary files (images, archives and so on)

2. Manage conversation history

Command	When to use	Effect
`/clear`	Before a new task	Clears all history and resets the context

3. Give precise instructions

❌ Vague	✅ Precise
"Optimize this code"	"Refactor `parse_data` in utils.py to use a list comprehension"
"Fix this for me"	"Add error handling at line 45 of index.js to catch API failures"
"Something's wrong"	"Fix the division-by-zero in calc.py and add input validation"

FAQ

Q1: How do I switch to a specific model?

Type /models and search by name (such as claude or gpt-5), then select the model. An empty result means the current version does not support that model yet.

Q2: Connection failures, 401 or 404 errors

Check the API key is correct (it starts with sk-)
Confirm the config path is ~/.config/opencode/opencode.json (the file is opencode.json, not config.json)
Confirm baseURL ends with /v1: https://api.atalk-ai.com/v1
Sign in to the SeaWhale AI console and check your balance
Quit and restart OpenCode after editing the config

Q3: Responses are slow

Switch to a lighter model (such as gpt-5-mini or claude-haiku-4-5-20251001)
Use /clear to drop a long conversation history and shorten the context

Q4: How do I update OpenCode?

bash

npm install -g opencode-ai@latest

Q5: Which programming languages are supported?

OpenCode supports all mainstream programming languages, including:

Supported languages

Web development: JavaScript, TypeScript, HTML, CSS, Vue, React

Backend development: Python, Java, Go, Rust, C/C++, C#, PHP, Ruby

Mobile development: Swift, Kotlin, Dart (Flutter)

Data science: Python (NumPy/Pandas), R, SQL

Other: Shell, YAML, JSON, Markdown

Error codes

HTTP status	Meaning	How to fix
401	API key auth failed	Check the API key; confirm the Base URL points at SeaWhale AI
403	Insufficient permission	Check API key permissions and model access
429	Rate limited	Wait a moment and retry, or increase your account quota
500	Server error	Retry later, or contact support

📚 Quick start — SeaWhale AI API basics
🎯 Model list — every available model
🔧 API reference — the complete API documentation

Other AI coding tools

Tool	Platform	Characteristics
OpenCode	CLI	Open source, lightweight, terminal-based
Claude Code	CLI	Official Anthropic tool, deep code understanding
Cline	VSCode	IDE extension, plans and runs multi-step tasks
Cherry Studio	Desktop app	Graphical interface, multi-model management

Support

Channel	Response time	Best for
📖 Documentation	Immediate	Common questions
💬 Live support	Weekdays 9:00–18:00	Real-time technical questions
📧 Email support	Within 24 hours	Detailed reports and feedback

🏠 Back to home
🔑 Get an API key

Last updated 2026-04-17

OpenCode integration guide ​

Introduction ​

Key capabilities ​

Why SeaWhale AI? ​

Supported models ​

Before you begin ​

1. Get a SeaWhale AI API key ​

2. System requirements ​

Installing OpenCode ​

Global install ​

Verify the installation ​

Configuring SeaWhale AI ​

Configuration file (recommended) ​

Getting started ​

Starting OpenCode ​

Common commands ​

Switching models ​

Worked examples ​

Example 1: code generation ​

Example 2: debugging ​

Saving tokens ​

1. Start in the specific project directory ​

2. Manage conversation history ​

3. Give precise instructions ​

FAQ ​

Q1: How do I switch to a specific model? ​

Q2: Connection failures, 401 or 404 errors ​

Q3: Responses are slow ​

Q4: How do I update OpenCode? ​

Q5: Which programming languages are supported? ​

Error codes ​

Related resources ​

Other AI coding tools ​

Support ​

OpenCode integration guide

Introduction

Key capabilities

Why SeaWhale AI?

Supported models

Before you begin

1. Get a SeaWhale AI API key

2. System requirements

Installing OpenCode

Global install

Verify the installation

Configuring SeaWhale AI

Configuration file (recommended)

Getting started

Starting OpenCode

Common commands

Switching models

Worked examples

Example 1: code generation

Example 2: debugging

Saving tokens

1. Start in the specific project directory

2. Manage conversation history

3. Give precise instructions

FAQ

Q1: How do I switch to a specific model?

Q2: Connection failures, 401 or 404 errors

Q3: Responses are slow

Q4: How do I update OpenCode?

Q5: Which programming languages are supported?

Error codes

Related resources

Other AI coding tools

Support