mcp client
MCP clients
Primary integration path for Claude Code, Cursor, and other MCP clients.
npx -y @nanogpt/mcp with NANOGPT_API_KEY in the MCP server environment.
Open docsMCP-first agent tooling
NanoGPT Agent Harness Pack
One place for NanoGPT agent endpoints, MCP setup, model catalogs, media tools, search, scrapers, and harness adapter instructions.
Harness Adapters
skill
OpenClaw
Use NanoGPT as an OpenAI-compatible model provider and expose NanoGPT tools through MCP.
Configure baseUrl https://nano-gpt.com/api and add the NanoGPT MCP server to OpenClaw tool configuration.
Open docstoolset
Hermes
Use NanoGPT as the model backend and load the NanoGPT MCP server as a dynamic MCP toolset.
Set the OpenAI-compatible base URL to https://nano-gpt.com/api and add @nanogpt/mcp as an MCP server.
openai compatible
OpenAI-compatible clients
Use NanoGPT directly from clients that support OpenAI chat, responses, images, audio, embeddings, or models APIs.
Set base URL to https://nano-gpt.com/api and Authorization: Bearer $NANOGPT_API_KEY.
Open docsEndpoint Matrix
MCP is the default agent surface. API-only routes are still usable directly from OpenAI-compatible clients.
Chat and Reasoning
Model calls for agent planning, synthesis, tool calling, and streaming.
| Capability | Endpoint | Auth | Mode | Inputs | Output | MCP |
|---|---|---|---|---|---|---|
| Chat Completions OpenAI-compatible chat, streaming, reasoning controls, tool calling, provider routing, and web-enabled model suffixes. | POST /api/v1/chat/completions | API key | Streaming | model, messages, stream?, tools?, tool_choice?, reasoning?, provider? | OpenAI chat completion or SSE stream | MCP readynanogpt_chat |
| Responses API OpenAI Responses-compatible endpoint for newer agent clients and stored response workflows. | POST /api/v1/responses | API key | Async or sync | model, input, stream?, tools?, previous_response_id?, metadata? | Responses object or SSE stream | MCP nextnanogpt_responses |
| Anthropic Messages Anthropic-compatible messages endpoint for clients that expect Claude-style request and response shapes. | POST /api/v1/messages | API key | Streaming | model, messages, max_tokens?, tools?, tool_choice?, stream? | Anthropic message object or event stream | API only |
Model Catalogs
Discover model IDs, pricing, modalities, and supported parameters.
| Capability | Endpoint | Auth | Mode | Inputs | Output | MCP |
|---|---|---|---|---|---|---|
| Text Model Catalog List chat/text models, capabilities, context windows, pricing, subscription inclusion, and provider options. | GET /api/v1/models | Public | Sync | detailed? | OpenAI-compatible model list | MCP readynanogpt_list_text_models |
| Image Model Catalog List image generation/editing models, modalities, pricing, resolutions, and supported parameters. | GET /api/v1/image-models | Public | Sync | detailed? | Model list | MCP readynanogpt_list_image_models |
| Video Model Catalog List video models, text/image/video input support, pricing, durations, resolutions, and provider-specific settings. | GET /api/v1/video-models | Public | Sync | detailed? | Model list | MCP readynanogpt_list_video_models |
| Audio Model Catalog List TTS, STT, voice, and music-capable audio models with pricing and supported parameters. | GET /api/v1/audio-models | Public | Sync | detailed?, type? | Model list | MCP readynanogpt_list_audio_models |
Search and Fetch
Fresh information retrieval and URL transcript/content extraction.
| Capability | Endpoint | Auth | Mode | Inputs | Output | MCP |
|---|---|---|---|---|---|---|
| Web Search Search the web through NanoGPT providers, including lightweight and deeper semantic modes. | POST /api/web | API key or session | Sync | query, provider?, depth?, outputType?, limit? | Search result JSON | MCP readynanogpt_web_search |
| URL Scraping Fetch and extract page content from public URLs, with optional stealth mode where supported. | POST /api/scrape-urls | API key or session | Sync | urls, stealthMode?, includeRawHtml? | Per-URL scrape result JSON | MCP readynanogpt_scrape_urls |
| YouTube Transcripts Retrieve YouTube transcripts for agent research and summarization workflows. | POST /api/youtube-transcribe | API key or session | Sync | urls | Transcript result JSON | MCP readynanogpt_youtube_transcribe |
Structured Scrapers
Application-grade social, maps, website, and profile data tools.
| Capability | Endpoint | Auth | Mode | Inputs | Output | MCP |
|---|---|---|---|---|---|---|
| Website Crawler Scrape, map, or crawl websites through the public Firecrawl-backed API route. | POST /api/v1/firecrawl | API key or session | Async or sync | operation, url?, urls?, limit?, formats?, maxDepth? | Crawler result JSON | MCP readynanogpt_firecrawl |
| Reddit Scraper Scrape Reddit posts, comments, communities, users, and search results. | POST /api/v1/reddit | API key or session | Async job | startUrls?, searches?, maxItems?, maxComments?, searchTypes? | Reddit scrape result JSON | MCP readynanogpt_reddit |
| X Research Retrieve public X search, posts, users, timelines, lists, Spaces, and counts. | GET /api/v1/x | API key or session | Sync | endpoint, query?, id?, ids?, username?, limit?, next_token? | X API style JSON | MCP readynanogpt_x_research |
| Google Maps Scraper Scrape Google Maps places, details, contacts, reviews, images, and enrichment fields. | POST /api/v1/googlemaps | API key or session | Async job | searchStringsArray?, startUrls?, placeIds?, resultLimit?, maxReviews?, scrapeContacts? | Place result JSON | MCP readynanogpt_google_maps_scrape |
| Google Maps Reviews Scrape Google Maps reviews, reviewer metadata, ratings, dates, and localized review text. | POST /api/v1/googlemaps/reviews | API key or session | Async job | startUrls?, placeIds?, maxReviews?, reviewsSort?, reviewsStartDate? | Review result JSON | MCP readynanogpt_google_maps_reviews_scrape |
| TikTok Scraper Scrape hashtags, profiles, searches, video URLs, comments, subtitles, media, and metadata. | POST /api/v1/tiktok | API key or session | Async job | hashtags?, profiles?, searchQueries?, postURLs?, resultsPerPage?, commentsPerPost? | TikTok scrape result JSON | MCP readynanogpt_tiktok_scrape |
| Instagram Profile Scraper Scrape public Instagram profile metadata and optional about-section data. | POST /api/v1/instagram/profile | API key or session | Async job | usernames, includeAboutSection?, resultLimit? | Instagram profile result JSON | MCP readynanogpt_instagram_profile |
| Instagram Posts Scraper Scrape Instagram posts, reels, captions, media metadata, and engagement data. | POST /api/v1/instagram/posts | API key or session | Async job | directUrls?, username?, resultsLimit?, search?, addParentData? | Instagram post result JSON | MCP readynanogpt_instagram_posts_scrape |
| Instagram Reels Scraper Scrape Instagram reels, captions, media metadata, and engagement fields. | POST /api/v1/instagram/reels | API key or session | Async job | directUrls?, username?, resultsLimit?, search? | Instagram reels result JSON | MCP readynanogpt_instagram_reels |
| Facebook Posts Scraper Scrape public Facebook page posts, captions, reactions, links, media, and metadata. | POST /api/v1/facebook/posts | API key or session | Async job | startUrls?, resultsLimit?, dateFrom?, dateTo? | Facebook post result JSON | MCP readynanogpt_facebook_posts_scrape |
| Facebook Ads Library Scraper Scrape Facebook Ads Library entries, campaign metadata, creative details, and advertiser pages. | POST /api/v1/facebook/ads | API key or session | Async job | startUrls?, resultsLimit?, activeStatus?, onlyTotal?, isDetailsPerAd? | Facebook ads result JSON | MCP readynanogpt_facebook_ads |
| LinkedIn Profile Scraper Scrape public LinkedIn profile data through NanoGPT billing. | POST /api/v1/linkedin/profile | API key or session | Async job | profileUrls?, usernames?, resultLimit? | LinkedIn profile result JSON | MCP readynanogpt_linkedin_profile |
Image and Video
Image generation/editing and async video workflows.
| Capability | Endpoint | Auth | Mode | Inputs | Output | MCP |
|---|---|---|---|---|---|---|
| Image Generation Generate images from text or image inputs with OpenAI-compatible response shapes. | POST /api/v1/images/generations | API key or session | Async or sync | model?, prompt, n?, size?, resolution?, aspect_ratio?, response_format? | Image URLs or base64 image data | MCP readynanogpt_image_generate |
| Image Editing Edit, inpaint, or transform images with image editing models. | POST /api/v1/images/edits | API key or session | Async or sync | model?, prompt, imageDataUrl?, imageDataUrls?, maskDataUrl?, n?, size? | Image URLs or base64 image data | MCP readynanogpt_image_edit |
| Video Generation Generate, extend, edit, or upscale video through async model-specific media jobs. | POST /api/generate-videosubmit | API key or session | Async job | model, prompt?, imageUrl?, imageDataUrl?, videoUrl?, duration?, resolution?, aspect_ratio? | Video job submission JSON with status lookup | MCP readynanogpt_video_generate |
| Video Status Unified generated video status lookup using request or run identifiers. | GET /api/video/statusstatus | API key or session | Sync | requestId?, runId?, modelSlug? | Video job status JSON | MCP readynanogpt_video_generate |
Audio and Music
Speech, transcription, voice, and music-capable generation.
| Capability | Endpoint | Auth | Mode | Inputs | Output | MCP |
|---|---|---|---|---|---|---|
| Text to Speech Generate speech audio from text with OpenAI-compatible TTS transport. | POST /api/v1/audio/speech | API key or session | Sync | model?, input, voice?, response_format?, speed?, instructions? | Audio bytes or JSON metadata | MCP readynanogpt_text_to_speech |
| Music Generation Generate music through music-capable audio models exposed on the audio speech transport, using a first-class agent wrapper. | POST /api/v1/audio/speech | API key or session | Async or sync | model, input or text, duration?, lyrics?, tags?, audio?, sampleRate?, bitrate? | Audio bytes or JSON metadata | MCP readynanogpt_music_generate |
| Speech to Text Transcribe audio files or URLs with OpenAI-compatible transcription inputs. | POST /api/v1/audio/transcriptions | API key or session | Sync | file or file_url, model?, language?, prompt?, response_format?, timestamp_granularities? | Transcription JSON or text | MCP readynanogpt_audio_transcribe |
Data Enrichment
Lead, company, email, and public data enrichment APIs.
| Capability | Endpoint | Auth | Mode | Inputs | Output | MCP |
|---|---|---|---|---|---|---|
| Hunter Email Data Find and verify professional emails, enrich people and companies, and inspect domain email counts. | GET /api/v1/hunter | API key or session | Sync | endpoint, domain?, company?, email?, full_name?, first_name?, last_name? | Hunter API style JSON | MCP readynanogpt_hunter |
Extras
Embeddings, detection, and other agent support utilities.
| Capability | Endpoint | Auth | Mode | Inputs | Output | MCP |
|---|---|---|---|---|---|---|
| Embeddings Create vector embeddings for text or token inputs. | POST /api/v1/embeddings | API key | Sync | model, input, encoding_format?, dimensions? | OpenAI-compatible embedding response | MCP readynanogpt_embeddings |
| AI Detection Detect AI-written text or plagiarism through NanoGPT AI Detection. | POST /api/v1/ai-detection | API key or session | Sync | text, mode? | Detection result JSON | MCP readynanogpt_ai_detection |