RouterPlex

One API key. 25+ AI models. Every major provider.
Pay per token. No markup. No subscriptions.
$5 free credit →

[RouterPlex @RouterPlex]

JavaScript:

import OpenAI from "openai";

const client = new OpenAI({
baseURL: "https://t.co/mvRbIg86zz",
apiKey: "YOUR_ROUTERPLEX_KEY" });

[RouterPlex @RouterPlex]

How to point an app that uses the OpenAI SDK at RouterPlex.

Two lines. Five minutes. Thread.

[RouterPlex @RouterPlex]

Python:

from openai import OpenAI

client = OpenAI(
base_url="https://t.co/mvRbIg7yK1",
api_key="YOUR_ROUTERPLEX_KEY")

[RouterPlex @RouterPlex]

https://t.co/stBI7iQwNJ

[RouterPlex @RouterPlex]

$250 → $500 of credit, then a hard limit per key.

Give each agent its own key with a $20 limit. When one loops, that key returns 402 at $20 and the other $480 is untouched. Nothing goes negative, nothing to pay off.

[RouterPlex @RouterPlex]

How keys and limits work:
https://t.co/NY2RESeLQa

[RouterPlex @RouterPlex]

Setup for Claude Code, OpenCode and the SDK:
https://t.co/L1XcC83QgG

[RouterPlex @RouterPlex]

Kimi K3.1 Pro

OpenRouter: $2 / $10 + 5.5% on every top-up
RouterPlex: $2 / $10 + $0, and $25 loads as $50

Same list price. 5M output tokens of Kimi for $50. Which cost you $25.

[RouterPlex @RouterPlex]

Sign up and top up here: https://t.co/tpo4WzzoXS

[RouterPlex @RouterPlex]

$25 of API credit.

OpenRouter: $25 of credit costs $26.38 (5.5% top-up fee).
RouterPlex: $25 buys $50 of credit. $0 fee.

Same vendor list prices on the models both of us route. The difference is what happens to the money before it reaches a token.

[RouterPlex @RouterPlex]

OpenRouter's fee is 5.5% with a $0.80 minimum, from their pricing page. Ours is $0. Fee math in full:
https://t.co/Tf9RTUAE6i

[RouterPlex @RouterPlex]

The match is applied on the billing page when you top up. Match credit works on the open-source models, not the abliterated ones.
https://t.co/X7u2y6oqq4

[RouterPlex @RouterPlex]

Top-up match on RouterPlex.

Load $25 → $50 of credit.
Load $250 → $500 of credit.

One match per customer. $0 top-up fee. Credit never expires. Balance stops at $0, never goes negative.
One key, ~40 open-source models at the vendor's list price.

[RouterPlex @RouterPlex]

The route just got stronger.
RouterPlex is now in @AWSstartups .

One API key → global-scale AI infrastructure.
#AWSActivate #AIInfrastructure #Startups #RouterPlex

[RouterPlex @RouterPlex]

Kimi K3's 1M-token context is useful for coding agents—but token limits are not spend limits.

At Moonshot's published rates:
• cached input: $0.30/M
• fresh input: $3/M
• output: $15/M

Give each agent key a hard dollar cap before long runs.

[RouterPlex @RouterPlex]

@KimiDevs @cline Copy the exact Cline fields and verify the first request:

https://t.co/2FGTLQ1LNQ

[RouterPlex @RouterPlex]

Using Kimi K3 from @KimiDevs in @cline is a configuration change, not an SDK rewrite.

Choose OpenAI Compatible, enter the RouterPlex /v1 endpoint, and use model ID kimi-k3.

Give the Cline key a hard budget before letting an agent run.

[RouterPlex @RouterPlex]

Full Kimi K3 pricing breakdown and OpenAI SDK setup:

https://t.co/5NZfg0MScJ

[RouterPlex @RouterPlex]

Moonshot's published Kimi K3 API rates:

Cached input: $0.30 / 1M
Input: $3 / 1M
Output: $15 / 1M

1,048,576-token context. Vision. OpenAI-compatible.

RouterPlex uses the same rates under model ID kimi-k3.

[Cline @cline]

Try in Cline with model id: moonshotai/kimi-k3

To install Cline in your CLI: npm i -g cline

Available as an extension for VS Code and JetBrains as well!

(Free, open source, bring your own API key and use any model)

https://t.co/2PULMWjvkz