Connect Once and Everywhere
Search, scraping, browser-automation, voice and model-inference providers behind a single key — native endpoints, one credit balance, usage at each provider's own price, and benchmark scores published with their run dates.
curl https://api.nativeport.ai/serper/search \ -H "Authorization: Bearer $NATIVEPORT_API_KEY" \ -H "Content-Type: application/json" \ -d '{"q": "your query"}' # → HTTP 200 · Serper's JSON, unmodified # balance at $0 → HTTP 402, never silent
Who leads each capability
Every provider runs the same task corpus per capability. These are the current front-runners — click through for the full ranked table, including the providers that scored poorly.
32 providers, one key
Native endpoints, native parameters, native response shapes — NativePort injects the upstream credential server-side and debits your balance at the provider's real cost. Nothing to rewrite when you arrive, nothing to unwind if you leave.
Search
11 providers
A 30B+ page search index built independently, spanning news, video, image and web verticals.
SERP and SEO intelligence — keywords, rankings, backlinks — with the deepest taxonomy of SERP elements.
Semantic search driven by neural embeddings, offering find-similar, clean extraction and an integrated answer endpoint.
A URL-to-markdown Reader that benchmarks at the top of its class, plus search that hands back full-page markdown in a single call.
A search alternative that stays ToS-clean, offering sourced answers natively, structured JSON output, and markdown via /fetch.
An agent-optimized index where the ranking follows a natural-language objective; results arrive as pre-compressed excerpts.
One endpoint serving search results from 80-plus engines, Google, Bing and YouTube among them.
Google plus 80+ additional search engines as structured results — key handling is taken off your plate.
Google search results across web, news, images, places, maps and scholar verticals, plus an endpoint for scraping webpages.
A search index native to agents and RAG, returning snippets with relevance scores and an optional answer.
Own-index web search with an accompanying Research API that cites sources, aimed at grounding LLMs.
Scraping & Crawling
11 providers
Web Unlocker for anti-bot targets, pre-built structured scrapers for hundreds of domains, and a documented compliance posture.
Structured scrapers and markdown output alongside a managed asynchronous crawl queue; billing follows success, scaled by domain complexity.
LLM-ready markdown that tops the benchmarks, plus schema-driven extraction, asynchronous crawling and an official MCP.
Turns a PDF or an image into page-level Markdown in one call, with the exact fields you name pulled out by JSON schema in the same pass.
Enterprise scraping backed by a proxy network in 195-plus countries, dedicated parsers, and success-only billing.
General scraping behind a single GET, with proxy rotation and automatic retries handled.
Strong anti-bot bypass behind a single flag (asp=true), markdown and AI output, and a per-response cost breakdown.
A single REST surface: JS rendering, residential and stealth proxies, extraction via AI or CSS, markdown out.
Recursive crawling at speed, markdown that comes back clean, and per-page USD costs reported inside the response.
One key spans a scraper API together with a CDP scraping browser; heavy anti-bot sits behind a lone flag.
A single endpoint pairing automatic ban-management with AI extraction typed for product, article and job pages.
Browser Automation
4 providers
Scraping and automation via a 7,000+ Actor marketplace, consolidated on one key and one bill.
Cloud headless browser sessions, driven through Playwright or Puppeteer, suited to auth-walled and multi-step flows.
One call in, a rendered result out: screenshot, PDF, HTML, or a CSS-selector scrape from a single page load.
Three fixed actions — scrape to Markdown/HTML, screenshot, or PDF — one page load in, one response out.
Voice & Audio
2 providers
Wildcard routing exposes the complete ElevenLabs REST surface for voice and text-to-speech.
Four fixed routes reach Fish Audio's text-to-speech, transcription, and from-a-description voice-design models, each metered on the exact unit Fish Audio itself bills — bytes, audio-hours, or a flat per-call rate.
Model Inference
4 providers
Claude's Messages API behind the same policed-route pattern as the OpenAI proxy, cache-tier pricing included.
xAI's Grok chat and Responses API, billed from the dollar cost each response reports rather than a maintained price table.
Hugging Face's Inference Providers router, fanned out across competing serving providers behind one chat-completions endpoint.
OpenAI's Responses, Chat Completions, embeddings and moderation endpoints, metered per token against the same balance as the rest of the catalog, with gpt-5.6's cache- and context-tiered Standard rates read straight from each response.
Picking model calls apart from routing policy? See how this inference route compares to OpenRouter or Eden AI.
Three moves, no migration
1 · Create an account
Sign-up mints your first key automatically and seeds the balance with $5 of credit — card comes later, if ever.
2 · Swap the base URL
Keep the provider's own SDK or raw HTTP calls. Point them at api.nativeport.ai/<provider>/… and authenticate with your NativePort key.
3 · Watch the ledger
Each request lands in your ledger with the provider's exact metered cost against it. One balance covers the whole catalog.
Built to be left alone
Agents run unattended; the gateway's behavior has to be predictable enough to build on. These are the load-bearing guarantees.
Upstream keys never travel
Provider credentials are injected per request on our side. The only secret in your agent's environment is its NativePort key.
Failure you can branch on
A drained balance answers HTTP 402 — never a quietly degraded response. Optional auto-top-up keeps long runs funded.
Rankings with receipts
Quality, latency p50, the track-specific cost metric and error rate, measured on shared corpora and published with run dates. Weak scores stay on the board.
See the boards →Readable by machines
The catalog and every scorecard ship as llms.txt, llms-full.txt and evals.json for agents that choose their own tools.