TalkCodyTalkCody

TalkCody Free Use Guide

Learn how to use various powerful AI models for free in TalkCody

TalkCody is designed to make AI accessible to every developer to enhance efficiency. If you don't want to purchase additional API credits, you can use TalkCody for free through several methods.

Free Plans Overview

Currently, TalkCody supports the following major free/low-cost options:

ProviderTypeAdvantagesLimitations
TalkCody FreeCompletely FreeBuilt-in zero config, supports high-perf modelsRequires GitHub login
Qwen CodeCompletely FreeAlibaba's Qwen model, strong coding capabilityRequires OAuth authentication
Google AI StudioPersonal Free TierGemini 2.5 series, ultra-long context1500 requests per day limit
GitHub CopilotSubscription ReuseReuse existing subscription, no extra costRequires a GitHub Copilot subscription
OllamaLocal RunningCompletely free, offline privacy, unlimited useDepends on local hardware performance
Ollama CloudCloud Free TierAccess top closed-source models (MiniMax, GLM, etc.)Daily/Weekly quota limits
LM StudioLocal RunningUser-friendly UI, OpenAI API compatibleDepends on local hardware performance
OpenRouterAggregated Free ModelsWide variety, early access to new modelsDepends on OpenRouter platform

1. TalkCody Free

TalkCody's official free service, designed to lower the barrier to AI entry and let users start immediately.

GitHub Login

To prevent API abuse, using TalkCody Free requires logging in with your GitHub account.

Select Model

In the TalkCody model selector, switch to the TalkCody Free provider.

Free access to MiniMax M2.1

Currently, TalkCody Free provides long-term free access to the MiniMax M2.1 model. This model has excellent performance in both general conversation and coding tasks.


2. Qwen Code

TalkCody supports calling Alibaba's Qwen coding models for free via Qwen Code OAuth.

Log in to Qwen Code

Log in to the Qwen Code client first.

Enter Qwen Code Token Path

In TalkCody Settings -> API Keys, paste the Token path for Qwen Code into the corresponding input field. TalkCody automatically detects the Qwen Code client's Token file location on each platform, or you can manually specify the path.

Configure Models

In Model Settings, switch the primary or secondary model to Qwen3 Coder Plus or Qwen3 Coder Flash to start using it.


3. Google AI Studio (Gemini)

Google provides a very generous free tier for developers.

Free Tier Details: According to Google's latest API policy, Gemini 2.5 Flash / Flash Lite models support 15 requests per minute (RPM) and 1500 requests per day (RPD) in the free tier.

Get API Key

Visit Google AI Studio and log in with your Google account.

Create Key

Click "Create API key" and copy the generated key.

Configure in TalkCody

In TalkCody Settings -> API Keys, paste the key into the Google AI input field.


4. GitHub Copilot Subscription Reuse

GitHub Copilot has subscription plans with free quotas, and TalkCody supports reusing these subscriptions to use Copilot models for free.

Please refer to the GitHub Copilot Guide for detailed configuration steps.


5. Ollama (Local LLM)

If you have a computer with decent specifications (16GB+ RAM recommended), you can run models locally.

Install Ollama

Visit ollama.com to download and install the client for your system.

Download Models

Run the following commands in your terminal to download models suitable for coding:

ollama pull qwen2.5-coder
# Or use a high-performance small model
ollama pull llama3.2

Automatic Connection

As long as the Ollama service is running in the background, TalkCody will automatically detect it. You will see Ollama models in the dropdown menu of Model Settings.

The biggest advantages of running models locally are privacy/security and being completely free.


6. Ollama Cloud (Cloud Free Models)

Ollama not only supports local execution but also offers the Ollama Cloud service, allowing users to call high-performance cloud models directly through the local Ollama client.

Free Quota: Ollama Cloud provides users with daily and weekly free usage quotas. It is an excellent way to experience top-tier domestic and international models (like MiniMax, GLM, Gemini, etc.) without configuring complex API keys.

Supported Models List

Through Ollama Cloud, you can access the following models for free:

  • MiniMax M2.1: minimax-m2.1:cloud (High-performance model)
  • GLM 4.7: glm-4.7:cloud (Latest model from Zhipu)
  • Gemini 3 Flash: gemini-3-flash-preview:cloud (Google's high-speed model)
  • Kimi K2 Thinking: kimi-k2-thinking:cloud (Moonshot's reasoning model)

Usage Steps

Ensure Ollama is Installed

If you haven't installed it yet, visit ollama.com.

Register/Login to Ollama Account

Execute the following command in your terminal to log in (if not already logged in):

ollama login

Run Cloud Models

You can directly select models with the :cloud suffix in the TalkCody model dropdown menu.

If you encounter a Model not found error during use, please run the corresponding command in your terminal once to activate the model:

# Activate MiniMax
ollama run minimax-m2.1:cloud

# Activate GLM 4.7
ollama run glm-4.7:cloud

# Activate Gemini 3 Flash
ollama run gemini-3-flash-preview:cloud

Use in TalkCody

As long as the Ollama service is running, TalkCody will automatically sync these cloud models. You can switch to them directly in Model Settings.


7. LM Studio (Local LLM)

LM Studio is another popular desktop application for running open-source LLMs locally, providing an intuitive interface to discover, download, and run models.

Install LM Studio

Visit lmstudio.ai to download and install the client for your system.

Download and Load Models

Search for and download the models you need (such as Qwen2.5-Coder or Llama 3.2) in LM Studio, then click "Load Model" to load it into memory.

Start Local Server

Click the Local Server icon (double arrow symbol) in the left sidebar of LM Studio, then click Start Server. The default port is usually 1234.

Configure in TalkCody

In TalkCody Settings -> API Keys, enable the LM Studio switch. TalkCody will automatically detect the local service running on port 1234, and you can select the loaded models in Model Settings.


8. OpenRouter

OpenRouter is a powerful model aggregation platform that not only allows you to access various models through a unified interface but also provides a wealth of free resources.

Key Advantages

  • Rich Variety of Free Models: OpenRouter consistently maintains a large number of free models. You can visit their model list and filter for "Free" to see what's currently available.
  • Early Access to New Models: Many AI vendors choose to provide limited-time free testing quotas on OpenRouter when launching new models for public beta, making it an excellent way to experience cutting-edge technology.

Configuration Steps

Get API Key

Visit OpenRouter to register and generate an API Key.

Configure in TalkCody

In TalkCody Settings -> API Keys, paste the key into the OpenRouter input field.

Discover Free Models

In the Model Settings, you can search for and add free models provided by OpenRouter (usually with a :free suffix).