Editorial Policy
Updated
BroadRouter is BroadNet's own site and links to its paid API, so we disclose that interest and follow fixed rules: BroadNet prices come only from its published price file, other providers' prices from their own pages with a source, date and quote, outside facts are cited and dated, we say when competitors are cheaper or better, and errors are corrected.
Who we are
BroadRouter.com is a developer guide to BroadNet Inference, the OpenAI-compatible LLM and speech-to-text API from Broadnet Technologies LLC, New Jersey, United States. The content is produced by the BroadNet Inference team; about BroadRouter covers who runs the site and how to reach us.
Our commercial interest
This site is not independent. It exists to help developers and, where it fits, to send them to BroadNet Inference; sign-up links carry campaign tags so BroadNet can see which pages help. We still write for the reader: articles say when another provider is cheaper per token, when a different tool suits a job better, and when BroadNet is not the right choice.
How articles are written and updated
- Articles are drafted with AI assistance and checked against the sources listed at the end of each page before publication.
- Every article opens with a direct answer and lists its sources. Facts about other companies link to their own pages, with the date we checked them.
- If we cannot verify a claim, we leave it out or say it is unconfirmed. We do not publish invented benchmarks, statistics or testimonials.
- Pages show when they were published and last updated. We update them when prices, limits or products change.
Prices and product facts
No BroadNet price is typed by hand. Prices and plan facts are inserted at build time from the published price file (last synced 2026-11-14), and the build fails if a page contains a hard-coded dollar amount. Worked examples are computed from the same file. If a page and BroadNet's live pricing disagree, the live pricing is correct.
How we collect third-party prices
- Source: each provider's own pricing page or price documentation, never a router, reseller or aggregator feed. 1,671 LLM API prices from 41 providers are kept in one file per provider in the site's code (
src/data/own-prices/). - Evidence on every row: the source URL, the date checked and a verbatim quote from the page. The row's notes say which region, tier or promotion the price applies to, and why a model was left out where a provider shows no public token price.
- Re-checks: every Monday a script re-fetches each source page and flags any row whose quote or price it can no longer find. It only reports; it never edits data. Each flagged page is then re-read and the value, quote and date updated, and the change is committed and shown in the price change log.
- Open data: every row, with its source, date and quote, is published as the LLM API provider price database (CSV and JSON).
These prices are read from the providers' pages with AI assistance and a page renderer, so they can contain mistakes; the quote on each row lets you check it. BroadNet's own prices are never typed by hand.
How we compare providers
Competitor prices, limits and policies come from their own public pages, are shown as list prices with the date checked, and can change at any time. Confirm with the provider before you buy. Claims BroadNet makes about itself, such as its Arabic speech-recognition leaderboard rank, are labelled as reported by BroadNet.
Tools and research data
The free tools run in your browser and state their formulas; the token counters use each model's official tokenizer pinned to an exact version. Our open datasets publish their method and inputs with a script to reproduce them, under CC BY 4.0.
Language
The site is in English. Pages about Arabic models and speech include Arabic examples.
Corrections
Found an error? Email info@broadnet.ai with the page address. We check it, fix confirmed errors, update the page's date and record the change in the public fact-check log.
Related: About · Terms of Use · Privacy policy · Refund policy.