Connecting Fluent Read

Difficulty 🖱️ Easy　|　API format NewAPI (built-in engine)　|　base_url https://byesu.com

Setup · fill these in

API addresshttps://byesu.com

API keysk-xxxxxxxxCreate in console →

Modelgpt-5.5

API format NewAPI · model names follow the console

Fluent Read is an open-source browser translation extension that supports 20-plus translation engines and AI models, giving you a native-level bilingual side-by-side reading experience. Point it at byesu — an AI API gateway with OpenAI-compatible and Anthropic-native endpoints — and one key lets you translate web pages with large models such as GPT, Claude, and Gemini, billed pay-as-you-go with no subscription.

Before you start

First go to Console → Tokens and create an API token (something like sk-xxxx). When you create it, be sure to pick the right group—the group determines which models you can call, and choosing the wrong one will cause a "no available channel" error. See Choosing a group for details.

Step 1: Install the extension

Fluent Read supports all major browsers, so install it on whichever one you use:

• Chrome: Search for "Fluent Read" in the Chrome Web Store.
• Edge: Search for "Fluent Read" in the Edge Add-ons store.
• Firefox: Search for "Fluent Read" in the Firefox Add-ons store.

After installation, the Fluent Read icon will appear in your browser toolbar.

Step 2: Configure byesu API

Click the Fluent Read icon in the toolbar to open the settings page. Fluent Read ships with a built-in NewAPI translation engine (byesu runs on the NewAPI architecture), so pick that one directly and fill in the fields as shown below:

Field	What to enter
Translation service	Choose NewAPI
NewAPI endpoint	https://byesu.com (without /v1—the extension completes the path automatically)
Access token	Your byesu token, something like sk-xxxx
Model	Pick one from the list, or choose "Custom model" and enter the model name below
Custom model	For example gpt-5.5

Recommended models

Translation tasks are sensitive to speed and cost-effectiveness, so choose based on your needs:

• gpt-5.5 — general-purpose translation, balanced speed and quality
• claude-opus-4-8 — better quality for long texts and professional documents
• gemini-3.1-pro-high — excellent performance with multiple languages and long context

The model name must match exactly the name listed under "Available models" in the console, and it must fall within the group selected for your token.

Don't get the API address wrong

With the NewAPI engine, the "NewAPI endpoint" must be https://byesu.com without /v1—the extension builds the path itself, so adding /v1 ends up as /v1/v1 and the connection fails. Only if you switch to the "OpenAI / Custom OpenAI compatible" engine does the API address need /v1 (https://byesu.com/v1). If neither works, check the "Stuck?" section below.

Fluent Read talks to byesu over the OpenAI-compatible endpoint (https://byesu.com/v1). byesu also exposes an Anthropic-native /v1/messages endpoint for clients like Claude Code — see Claude Code CLI Setup.

Step 3: Run it and verify

1. Save the settings.
2. Open any foreign-language web page (for example, an English article).
3. Click the Fluent Read icon and choose "Translate current page", or use its select-to-translate feature.
4. When the page shows a bilingual side-by-side translation, the connection is working.

If no translation appears after a while, first open the browser's right-click menu → Inspect → Console to see whether there are any errors, then check the troubleshooting steps below.

Stuck?

• 401 / invalid token: Check that the token was copied in full (including the sk- prefix) and has no extra spaces. See Common errors #auth.
• No available channel / no channel: This is usually because the wrong group was selected for the token, and that group does not contain the model you entered. Switch to the correct group, or use a model name that is within the group. See Choosing a group and Common errors #no-channel.
• Insufficient quota / not enough balance: The account or token has run out of quota. Go to the console to top up or adjust the token quota. See Common errors #balance.
• Request timeout / no translation: With the NewAPI engine, make sure the API address is https://byesu.com (without /v1); only with the OpenAI-compatible engine should it be https://byesu.com/v1. Don't enter some other placeholder address.

For more troubleshooting, see Common errors.