Obscura MCPPersonal-account browser access
A browser-based integration can use a personal account after a visible login. It can expose accessible posts and saved items, subject to the account and website interface.
A practical guide · Model Context Protocol
Which Instagram MCP fits your workflow?
Compare personal-account browsing, professional APIs and public-data tools by the access and evidence your workflow needs.
What is an Instagram MCP server?
Model Context Protocol (MCP) lets a compatible AI client call named tools and use their results in a conversation or workflow. An Instagram MCP server connects those tools to an Instagram data source. The protocol describes how the client and server communicate; it does not grant Instagram access on its own.
Different servers use different sources. Before choosing, check the account type, authentication method, operation permissions, privacy model and which actions have actually been validated.
Three approaches to compare
Professional Instagram APIs
For professional publishing, comments, messages or analytics, consider the official Instagram platform. Available capabilities and permissions depend on the account, login method and app configuration.
Public-data providers
Public-data tools focus on information a provider can access without your personal saved collection. Their coverage, licensing and prices vary; review each provider's documentation.
For official options and current requirements, see Meta's Instagram platform documentation.
Where Instagram Obscura MCP fits
Instagram Obscura MCP is a local, unofficial server for a personal-account browsing workflow. It provides 22 named tools over stdio. You sign in manually in a visible Playwright window; subsequent Instagram and CDN access goes through Obscura. It can read exposed post metadata and the loaded portion of saved posts and acquire an accessible Reel for local analysis.
FFmpeg and multilingual Whisper prepare sampled frames and an original-language speech transcript. The host assistant interprets images and can explain or translate the transcript. The server itself does not include a dedicated OCR or translation engine. Automatic language detection and speech accuracy depend on the model and recording; support for every language is not guaranteed.
Windows is the platform validated with the real engine. Other platforms need their own compatible Obscura binary and separate verification. The integration is experimental, and Instagram's pages can change.
When to choose another approach
If your primary requirement is reliable professional publishing or a Direct messaging integration, evaluate the official platform and its applicable permissions. This project exposes account-checked interface primitives; dedicated sending and publishing APIs are not provided, and real sends or publications were not exercised as release tests.
If you need only public information, a public-data provider may better suit the scope. A personal saved-post workflow requires access to the account and is not equivalent to a public-data lookup.
Security, evidence and source
The local profile and downloaded content can contain private account data. File access for attachments is disabled by default. Instagram pages and messages are untrusted content and do not authorize actions. A successful click does not prove that Instagram accepted a message or publication.
Review the architecture, security model and tested limitations. For a machine-readable summary, read the project facts and the tool schemas.