> ## Documentation Index > Fetch the complete documentation index at: https://docs.agipower.ai/llms.txt > Use this file to discover all available pages before exploring further. # Codex # Guide to Using Codex CLI with AGIPower Codex CLI is an open-source programming assistant from OpenAI that runs in your local terminal. It can read, modify, and run code within a directory you choose. Built with Rust, it is fast and efficient, and is continuously improved on GitHub. By integrating with AGIPower, you gain access to more model options instead of being limited to the official OpenAI API. OpenAI has explicitly positioned Responses as the next-generation unified interface. Chat Completions is still available, but it is no longer the preferred choice for new projects. Codex will follow this direction as well—treating Chat Completions as a compatibility option and gradually migrating to Responses (the configuration in this article uses Responses). Note the OpenAI-compatible setting: base\_url="[https://api.agipower.ai/v1](https://api.agipower.ai/v1)". ## Configuration Options ### Install Codex CLI ```bash title="npm/pnpm" theme={null} # Install with pnpm (recommended) pnpm install -g @openai/codex # Or install with npm npm install -g @openai/codex ``` ### Configure Environment Variables Add your AGIPower API Key to your shell configuration file: ```bash theme={null} # Edit ~/.zshrc or ~/.bashrc depending on the terminal you use export AGIPower_API_KEY="sk-xxx" # [!code highlight] ``` Make sure to replace `sk-xxx` with your actual AGIPower API Key. You can obtain an API Key in the [AGIPower Console](https://agipower.ai/settings/keys). ### Configure Codex Create or modify the Codex configuration file `~/.codex/config.toml`: ```toml theme={null} model_provider = "agipower" # [!code highlight] model = "openai/gpt-5.2-codex" # [!code highlight] [model_providers.agipower] # [!code highlight] name = "AGIPower" # [!code highlight] base_url = "https://api.agipower.ai/v1" # [!code highlight] env_key = "AGIPower_API_KEY" # [!code highlight] wire_api = "responses" # [!code highlight] ``` * `model_provider`: Use AGIPower as the model provider * `model`: The model to use; can be any model supported by AGIPower * `base_url`: The base URL of the AGIPower API * `env_key`: The name of the environment variable that stores the API Key * `wire_api`: Use the Responses protocol (recommended) ### Start Using It Immediately After configuration, reload your shell configuration and start Codex: ```bash theme={null} # Reload configuration source ~/.zshrc # or source ~/.bashrc # Enter your project directory cd my-project # Start Codex CLI codex # [!code highlight] ``` After adding the environment variable to your shell configuration file, you won’t need to set it manually each time. The configuration will apply automatically every time you open a new terminal. ## Supported Models You can freely switch the `model` field in `config.toml` to any model supported by AGIPower. * View models available under the Responses protocol via the [AGIPower model list](https://agipower.ai/models?sort=newest\&supported_protocol=responses) * Use the model slug (e.g., `openai/gpt-5.2-codex`) * To target a specific provider, see the [Provider Routing documentation](/en/guide/provider-routing) Below are some recommended models with strong coding performance: | # | Model slug | Highlights | | -- | ----------------------------- | ------------------------- | | 1 | `openai/gpt-5.2-codex` | Coding-optimized | | 2 | `openai/gpt-5.2` | Strong general capability | | 3 | `anthropic/claude-sonnet-4.5` | Excellent reasoning | | 4 | `anthropic/claude-opus-4.1` | Complex tasks | | 5 | `google/gemini-2.5-pro` | Multimodal support | | 6 | `x-ai/grok-code-fast-1` | Fast responses | | 7 | `x-ai/grok-4-fast` | Efficient coding | | 8 | `deepseek/deepseek-chat` | Cost-effective | | 9 | `qwen/qwen3-coder-plus` | Chinese-friendly coding | | 10 | `moonshotai/kimi-k2-0905` | Long-context support | | 11 | `z-ai/glm-4.6` | Balanced overall | | 12 | `inclusionai/ring-1t` | Strong reasoning | For more models, see the [AGIPower model list](https://agipower.ai/models?sort=newest\&supported_protocol=responses). ### Responses-Compatible Models (Fetched on 2026-01-29) The list below is fetched from the [AGIPower model list (Responses filter)](https://agipower.ai/models?sort=newest\&supported_protocol=responses). Use that link for the latest updates. #### Model slug list (Responses) ``` inclusionai/llada2.0-flash-cap z-ai/glm-4.7 minimax/minimax-m2.1 volcengine/doubao-seed-1.8 google/gemini-3-flash-preview google/gemini-3-flash-preview-free xiaomi/mimo-v2-flash xiaomi/mimo-v2-flash-free openai/gpt-5.2-codex openai/gpt-5.2-pro openai/gpt-5.2 openai/gpt-5.2-chat z-ai/glm-4.6v z-ai/glm-4.6v-flash-free z-ai/glm-4.6v-flash deepseek/deepseek-v3.2 mistralai/mistral-large-2512 deepseek/deepseek-chat deepseek/deepseek-reasoner anthropic/claude-opus-4.5 google/gemini-3-pro-image-preview x-ai/grok-4.1-fast x-ai/grok-4.1-fast-non-reasoning google/gemini-3-pro-preview openai/gpt-5.1 openai/gpt-5.1-chat openai/gpt-5.1-codex openai/gpt-5.1-codex-mini baidu/ernie-5.0-thinking-preview volcengine/doubao-seed-code moonshotai/kimi-k2-thinking moonshotai/kimi-k2-thinking-turbo qwen/qwen3-max-preview inclusionai/ming-flash-omni-preview minimax/minimax-m2 kuaishou/kat-coder-pro-v1 kuaishou/kat-coder-pro-v1-free anthropic/claude-haiku-4.5 inclusionai/ring-1t inclusionai/ling-1t google/gemini-2.5-flash-image openai/gpt-5-pro z-ai/glm-4.6 anthropic/claude-sonnet-4.5 deepseek/deepseek-v3.2-exp openai/gpt-5-codex qwen/qwen3-max qwen/qwen3-vl-plus x-ai/grok-4-fast x-ai/grok-4-fast-non-reasoning inclusionai/ling-flash-2.0 inclusionai/ring-flash-2.0 baidu/ernie-x1.1-preview moonshotai/kimi-k2-0905 inclusionai/ling-mini-2.0 inclusionai/ring-mini-2.0 x-ai/grok-code-fast-1 deepseek/deepseek-chat-v3.1 volcengine/doubao-seed-1-6-vision openai/gpt-5 openai/gpt-5-chat openai/gpt-5-mini openai/gpt-5-nano anthropic/claude-opus-4.1 stepfun/step-3 z-ai/glm-4.5 z-ai/glm-4.5-air qwen/qwen3-coder-plus google/gemini-2.5-flash-lite qwen/qwen3-235b-a22b-2507 qwen/qwen3-235b-a22b-thinking-2507 qwen/qwen3-coder moonshotai/kimi-k2-0711 x-ai/grok-4 google/gemini-2.5-flash google/gemini-2.5-pro deepseek/deepseek-r1-0528 anthropic/claude-opus-4 anthropic/claude-sonnet-4 google/gemma-3-12b-it qwen/qwen3-14b openai/o4-mini openai/gpt-4.1 openai/gpt-4.1-mini openai/gpt-4.1-nano meta/llama-4-scout-17b-16e-instruct google/gemini-2.0-flash-lite-001 anthropic/claude-3.7-sonnet google/gemini-2.0-flash meta/llama-3.3-70b-instruct anthropic/claude-3.5-haiku anthropic/claude-3.5-sonnet openai/gpt-4o-mini openai/gpt-4o ``` ## Troubleshooting ### Common Issues #### API Key Error **Issue**: You are told the API Key is invalid or unauthorized. **Solutions**: * Check whether the environment variable `AGIPower_API_KEY` is set correctly * Use `echo $AGIPower_API_KEY` to verify the environment variable value * Confirm the API Key is active and that you have sufficient balance * Verify the API Key format starts with `sk-` #### Connection Failure **Issue**: Codex CLI cannot connect to the AGIPower service. **Solutions**: * Check that your network connection is working * Verify that `base_url` in `config.toml` is correctly set to `https://api.agipower.ai/v1` * Confirm your firewall settings are not blocking outbound connections * Try `curl https://api.agipower.ai/v1/models` to test connectivity #### Environment Variable Not Taking Effect **Issue**: Even after setting the API Key, it still says it is not configured. **Solutions**: * Reopen the terminal window, or run `source ~/.zshrc` or `source ~/.bashrc` to reload the configuration * Confirm the environment variable is set correctly: `echo $AGIPower_API_KEY` * Make sure you added it to the correct shell configuration file (`.zshrc` for zsh, `.bashrc` for bash) #### Configuration File Path Issue **Issue**: Changes to the configuration file do not take effect. **Solutions**: * Confirm the configuration file path is `~/.codex/config.toml` * If the directory does not exist, create it first: `mkdir -p ~/.codex` * Check that the configuration file syntax is correct (TOML format) * Use `cat ~/.codex/config.toml` to verify the file contents #### Model Unavailable **Issue**: A model is reported as unavailable or unsupported. **Solutions**: * Visit the [AGIPower model list](https://agipower.ai/models?sort=newest\&supported_protocol=responses) to confirm the model is available * Check that the model slug is spelled correctly * Try another recommended model to test * Confirm your account has permission to access the model ## Advanced Configuration ### Configure Models of Different Sizes You can switch between different model sizes in `config.toml` based on your task requirements: ```toml title="Balanced Setup" theme={null} # Model choice that balances performance and cost model_provider = "agipower" model = "anthropic/claude-sonnet-4.5" [model_providers.agipower] name = "AGIPower" base_url = "https://api.agipower.ai/v1" env_key = "AGIPower_API_KEY" wire_api = "responses" ``` ```toml title="Performance-First Setup" theme={null} # Model choice optimized for performance model_provider = "agipower" model = "openai/gpt-5.2-codex" [model_providers.agipower] name = "AGIPower" base_url = "https://api.agipower.ai/v1" env_key = "AGIPower_API_KEY" wire_api = "responses" ``` ```toml title="Cost-Optimized Setup" theme={null} # Model choice optimized for cost efficiency model_provider = "agipower" model = "deepseek/deepseek-chat" [model_providers.agipower] name = "AGIPower" base_url = "https://api.agipower.ai/v1" env_key = "AGIPower_API_KEY" wire_api = "responses" ``` If you run into any issues during use, or if you have suggestions or feedback, feel free to contact us via: * **Official website**: [https://agipower.ai](https://agipower.ai) * **Technical support email**: [support@agipower.ai](mailto:support@agipower.ai) * **Business inquiries**: [support@agipower.ai](mailto:support@agipower.ai) * **Twitter**: [@AGIPowerAI](https://twitter.com/AGIPowerAI) * **Discord community**: [http://discord.gg/vHZZzj84Bm](http://discord.gg/vHZZzj84Bm) For more contact methods and details, please visit our [Contact page](/en/help/contact).