# BLACKBOX AI

## Docs

- [Authentication](https://docs.blackbox.ai/api-reference/authentication.md): Learn how to authenticate your requests to the BLACKBOX API.
- [API Best Practices](https://docs.blackbox.ai/api-reference/best-practices.md): How to correctly structure tool calls, tool results, and reasoning blocks in multi-turn conversations.
- [Create Chat Completion](https://docs.blackbox.ai/api-reference/chat.md): Send a chat completion request. Find your [API Key](https://app.blackbox.ai/dashboard).
- [Encrypted Model](https://docs.blackbox.ai/api-reference/encrypted-model.md): Send end-to-end encrypted messages to BLACKBOX AI models running inside a verified GPU enclave.
- [Errors](https://docs.blackbox.ai/api-reference/errors.md): API error handling
- [Introduction](https://docs.blackbox.ai/api-reference/introduction.md): Start building powerful applications with easy access to the state-of-the-art AI models.
- [Interleaved Thinking](https://docs.blackbox.ai/api-reference/messages/interleaved-thinking.md): Enable the model to reason between tool calls for better multi-step problem solving
- [Messages API](https://docs.blackbox.ai/api-reference/messages/introduction.md): Send requests using the Anthropic Messages API format for Claude models
- [Tool Calling](https://docs.blackbox.ai/api-reference/messages/tool-calling.md): Define tools the model can call and build multi-turn agentic loops with the Messages API
- [Model Alias](https://docs.blackbox.ai/api-reference/model-alias.md)
- [LLM Pricing](https://docs.blackbox.ai/api-reference/models/chat-pricing.md): Explore the AI models available for conversational tasks, text generation, and complex reasoning.
- [Parameters](https://docs.blackbox.ai/api-reference/parameters.md): Learn about all available parameters for BLACKBOX AI API requests. Configure temperature, max tokens, top_p, and other model-specific settings.
- [Prompt caching](https://docs.blackbox.ai/api-reference/prompt-caching.md)
- [Provider Routing](https://docs.blackbox.ai/api-reference/provider-routing.md): Control how your requests are routed across multiple AI providers. Optimize for cost, performance, or reliability with intelligent provider selection.
- [Reasoning and Interleaved Thinking](https://docs.blackbox.ai/api-reference/reasoning.md): Enable advanced reasoning capabilities and interleaved thinking for sophisticated AI responses with transparent reasoning steps.
- [Requests](https://docs.blackbox.ai/api-reference/requests.md): Learn about requests to BLACKBOX AI API
- [Responses](https://docs.blackbox.ai/api-reference/responses.md): Learn about response from BLACKBOX AI API
- [Reasoning](https://docs.blackbox.ai/api-reference/responses/reasoning.md): Enable reasoning models to think before they respond using the Responses API
- [Streaming](https://docs.blackbox.ai/api-reference/responses/streaming.md): Stream responses token by token using the Responses API for real-time output
- [Text Generation](https://docs.blackbox.ai/api-reference/responses/text-generation.md): Generate text responses using the Responses API with BLACKBOX AI models
- [Tool Calling](https://docs.blackbox.ai/api-reference/responses/tool-calling.md): Define tools the model can call to access external functions and APIs
- [Sample Usage](https://docs.blackbox.ai/api-reference/sample-usage.md): Examples of using BLACKBOX AI API for text models
- [Tool & Function Calling](https://docs.blackbox.ai/api-reference/tool-calling.md): Use tools in your prompts
- [Agent Runtimes — Claude, Codex & Grok Build](https://docs.blackbox.ai/api-reference/v1/agent-runtimes.md): How the Agent API runs your task on one of three fully-managed agent runtimes — the Claude Agent SDK, the OpenAI Codex SDK, or the Grok Build CLI — selected from the model id, and how all three behave identically from the caller's side.
- [Agent Run Status](https://docs.blackbox.ai/api-reference/v1/agent-status.md): Check the current status of a Claude Agent run by runId. Checks live in-memory state first, falls back to the database.
- [Agent SSE Stream](https://docs.blackbox.ai/api-reference/v1/agent-stream.md): Subscribe to the raw SSE event stream for a specific agent run by runId. Replays buffered events and streams new ones in real-time.
- [Authentication](https://docs.blackbox.ai/api-reference/v1/authentication.md): How to authenticate with the BLACKBOX Agent API using API keys and Bearer tokens.
- [Benchmark Logs & Status](https://docs.blackbox.ai/api-reference/v1/benchmark-logs.md): Poll a benchmark run's progress, read its orchestrator logs, or stream them live over SSE. Logs come from memory while the run is alive and from durable storage after it ends.
- [Benchmark Results](https://docs.blackbox.ai/api-reference/v1/benchmark-results.md): Final (or in-progress) score for a benchmark run — headline percentage, per-task pass/fail with verdict + duration, the underperformers, and an inline tracking log. Everything for one run, by id.
- [Run Benchmarks](https://docs.blackbox.ai/api-reference/v1/benchmarks.md): Launch a benchmark evaluation (SWE-bench, GAIA, AIME, GPQA, …) driven by the agent of your choice. Fire-and-forget: returns a benchmarkRunId immediately.
- [Cancel Task](https://docs.blackbox.ai/api-reference/v1/cancel-task.md): Cancel a running task or rename a task's chat title. Use action: 'cancel' to stop execution.
- [Continue Task](https://docs.blackbox.ai/api-reference/v1/continue-task.md): Send a follow-up prompt to an existing task, creating a new agent run on the same chat thread with full conversation history.
- [Create Task](https://docs.blackbox.ai/api-reference/v1/create-task.md): Create and execute a task using an AI agent. Supports both standard chat and Claude agent modes.
- [Get Workspace Diff](https://docs.blackbox.ai/api-reference/v1/files-diff.md): Get the cumulative git diff of all changes made by the agent in the sandbox workspace.
- [Download Workspace](https://docs.blackbox.ai/api-reference/v1/files-download.md): Download the entire sandbox workspace (or a subdirectory) as a zip or tar.gz archive.
- [List Workspace Files](https://docs.blackbox.ai/api-reference/v1/files-list.md): List files and directories in the task's sandbox workspace. Supports recursive listing.
- [Read / Write / Delete File](https://docs.blackbox.ai/api-reference/v1/files-read.md): Read, create/overwrite, or delete a single file in the task's sandbox workspace using GET, PUT, or DELETE.
- [Upload Files](https://docs.blackbox.ai/api-reference/v1/files-upload.md): Upload one or more files into the task's sandbox workspace using multipart/form-data.
- [Get Task](https://docs.blackbox.ai/api-reference/v1/get-task.md): Retrieve full details of a specific agent task run, including status, GitHub context, and the complete conversation message history.
- [List Branches](https://docs.blackbox.ai/api-reference/v1/git-branches.md): List branches for a GitHub repository. Optionally filter to protected branches only.
- [Get GitHub Config](https://docs.blackbox.ai/api-reference/v1/git-config-get.md): Returns the GitHub account linked to your BLACKBOX API key, or { connected: false } if no token is stored.
- [Connect Github](https://docs.blackbox.ai/api-reference/v1/git-config-set.md): Validate and store a GitHub personal access token. Enables repository access for agent tasks. Replaces any previously stored token.
- [Disconnect GitHub](https://docs.blackbox.ai/api-reference/v1/git-disconnect.md): Revoke and purge all stored GitHub tokens associated with your account. Clears both the database token and all GitHub-related cookies.
- [List Repositories](https://docs.blackbox.ai/api-reference/v1/git-repos.md): List all GitHub repositories accessible to your connected account, including personal and organization repositories.
- [GitHub Connection Status](https://docs.blackbox.ai/api-reference/v1/git-status.md): Check whether your GitHub account is connected and verify that your stored token is valid.
- [Introduction](https://docs.blackbox.ai/api-reference/v1/introduction.md): Programmatically create, monitor, and manage AI agent tasks using the BLACKBOX Agent API.
- [List Tasks](https://docs.blackbox.ai/api-reference/v1/list-tasks.md): Retrieve a paginated list of your agent task runs with optional status filtering.
- [Models](https://docs.blackbox.ai/api-reference/v1/models.md): Available models for the BLACKBOX Agent API. Use these model IDs in the `model` field when creating tasks.
- [Stream Task Logs](https://docs.blackbox.ai/api-reference/v1/stream-logs.md): Stream task execution logs in real-time using Server-Sent Events (SSE). Live runs stream from the in-memory buffer; completed runs replay from persisted message parts.
- [Task Lifecycle](https://docs.blackbox.ai/api-reference/v1/task-lifecycle.md): Understand how agent tasks move through states — from creation to completion — and how to track, stream, continue, or cancel them.
- [Get Task Logs](https://docs.blackbox.ai/api-reference/v1/task-logs.md): Retrieve the full execution log for a task as parsed stream events. For live tasks, events come from the in-memory buffer. For completed tasks, events are reconstructed from persisted message parts.
- [Get Task Status](https://docs.blackbox.ai/api-reference/v1/task-status.md): Lightweight status poll for a task run. Returns only status, progress, and timestamps — no messages or GitHub details.
- [Web Search](https://docs.blackbox.ai/api-reference/web-search.md): Enable real-time web search capabilities in your AI applications
- [Zero Data Retention](https://docs.blackbox.ai/api-reference/zdr.md)
- [ACP Integration with Zed Editor](https://docs.blackbox.ai/features/blackbox-cli/acp-integration.md): Complete guide to setting up Blackbox CLI with Zed editor using Agent Client Protocol (ACP)
- [Benchmarks](https://docs.blackbox.ai/features/blackbox-cli/benchmarks.md)
- [Chat With Your Database](https://docs.blackbox.ai/features/blackbox-cli/chat-with-database.md): Use BLACKBOX CLI to connect to your database and ask questions in plain English. Database tools run read-only queries and keep the connection active during your session for quick follow-ups.
- [Commands reference](https://docs.blackbox.ai/features/blackbox-cli/commands-reference.md)
- [Common use cases](https://docs.blackbox.ai/features/blackbox-cli/common-use-cases.md)
- [Getting started](https://docs.blackbox.ai/features/blackbox-cli/getting-started.md)
- [Introduction](https://docs.blackbox.ai/features/blackbox-cli/introduction.md)
- [Key features](https://docs.blackbox.ai/features/blackbox-cli/key-features.md)
- [Skills Management](https://docs.blackbox.ai/features/blackbox-cli/skills.md): Learn how to create and manage skills to extend your AI agent's capabilities with specialized knowledge
- [Best practices](https://docs.blackbox.ai/features/vscode-agent/best-practices.md)
- [Getting Started](https://docs.blackbox.ai/features/vscode-agent/getting-started.md): Step-by-step guide to install and set up the BLACKBOX AI VS Code extension.
- [Introduction](https://docs.blackbox.ai/features/vscode-agent/introduction.md): Learn about BLACKBOX AI VS Code extension and why it's the best choice for AI-powered development.
- [Key Features](https://docs.blackbox.ai/features/vscode-agent/key-features.md): Discover what you can accomplish with the BLACKBOX AI VS Code extension.
- [Orchestration](https://docs.blackbox.ai/features/vscode-agent/orchestration.md): Learn how to use Orchestration in the BLACKBOX AI VS Code Agent to run multiple sub-agents in parallel for efficient development of large features.
- [Voice Mode Powered by ElevenLabs](https://docs.blackbox.ai/features/vscode-agent/voice-mode.md): Interact with BLACKBOX AI VSCode Agent entirely through voice using ElevenLabs-powered speech-to-text and text-to-speech capabilities.

## OpenAPI Specs

- [openapi](https://docs.blackbox.ai/api-reference/openapi.json)