Frequently Asked Questions

MixRoute is an AI API gateway, often called an LLM gateway. You get one OpenAI-compatible API key that reaches 250+ models from OpenAI, Anthropic, Google, DeepSeek, xAI, Moonshot AI and more, and you move between them by changing the model name in your request.

Behind that single endpoint, MixRoute bills at each provider’s official price with 0% markup, sends your traffic with higher rate limits instead of through the public queue, and fails over automatically when a provider has trouble. It is built and operated by ELITE CLOUD PTE. LTD., a cloud service provider and official partner of AWS, GCP and Azure.

You can, and you would pay the same per-token price. The difference is everything around the tokens. Going direct means a separate account, key, invoice and dashboard for every provider, and your requests wait in the same public queue as everyone else’s.

MixRoute adds five things the official APIs don’t: one key for every provider, higher rate limits than going direct, automatic failover when a provider goes down, cross-timezone scheduling that keeps capacity working around the clock, and one bill instead of three to five separate accounts.

The same way cloud resellers have worked for decades. ELITE CLOUD is an authorized reseller for AWS, GCP and Azure with volume agreements, so MixRoute earns through provider partnerships and enterprise contracts rather than a fee on your tokens.

Model prices come straight from each provider’s published pricing page and are passed through 1:1. You can check any price on the models page against the provider’s own site, and every request is auditable line by line in your usage records.

Aggregation platforms typically add a 5–10% fee and route requests through the public queue; OpenRouter, for example, charges a 5.5% platform fee when you buy credits. MixRoute adds nothing to the official price, so your whole budget goes to tokens.

The bigger difference is infrastructure. MixRoute offers higher rate limits, schedules capacity across time zones, and fails over in milliseconds when a provider goes down, which is reliability a pure routing layer can’t offer. The API stays OpenAI-compatible, so moving over doesn’t mean rewriting your integration.

For most teams it’s two lines. Point your base URL at https://api.mixroute.ai/v1, swap in your MixRoute API key, and keep the SDK and request format you already use. Most teams are done in under five minutes.

After that, trying a different model is a one-parameter change: move a feature from GPT to Claude or DeepSeek by editing the model name, with no new account or integration. The quickstart walks through it step by step.

No. Prompts and responses pass through memory only: they come in, get routed, the answer goes back, and nothing is written, logged or used to train any model, ours included. MixRoute keeps only the metadata needed for billing and your usage dashboard: token counts, model, latency and cost.

If you need to control where traffic goes, Compliance Mode locks requests to whitelisted US and Singapore endpoints, and you can verify it at any time. The full details are in the data security FAQ and the Privacy Policy.

Higher rate limits run on dedicated throughput (Provisioned Throughput) that MixRoute buys from cloud providers. Your requests run through that reserved pool instead of the shared public queue every other user competes for, which means noticeably lower latency and near-zero 429 rate-limit errors, even at peak hours.

A global scheduling system shares the pool across time zones, so the capacity is always in use and nothing goes to waste. Teams with monthly usage of at least US$20,000 or a single top-up of at least US$30,000 can also discuss dedicated Reserved Capacity as part of an enterprise plan.

Yes. New accounts start at 50 RPM, rising to 100 RPM after the first top-up. Business plans support up to 6,000 RPM, and Enterprise plans can reserve dedicated capacity. If you need more at any tier, contact us with your expected peak RPM/TPM and models, and we’ll review it and confirm what we can provide.

MixRoute fails over for you. Traffic runs across multiple nodes, and when a provider has an outage the gateway switches in milliseconds, so your product keeps answering instead of passing errors on to your users.

You don’t have to build and maintain that fallback logic yourself, and you don’t need a second provider account on standby. A live status page is open to everyone, so you can see platform health at any time.

Smart Routing is completely free and optional. Create a Smart Routing API key in the console and use it in place of your current key; the request format stays the same. MixRoute’s in-house routing model reads each request in memory, judges its complexity and task type, and sends it to the best-fit model in the pool you authorize, so “When do you close?” isn’t billed at flagship rates. Across 10,000+ real-world runs the average saving was 20–40%.

MixRoute is pay as you go, and you don’t need a credit card to sign up. Self-serve top-ups accept Visa, Mastercard and JCB, plus USDT and USDC. Credits never expire, and larger top-ups earn bonus credit, from 2% on Growth up to 5% on Scale.

Business plans are prepaid and accept card, crypto, or bank wire transfer. Monthly usage of US$20,000 or more, or a single top-up of US$30,000 or more, qualifies for an enterprise plan: payment is mainly by bank wire transfer with an invoice and flexible payment terms, plus a custom contract and per-model pricing, with up to 25% off select models. Talk to Sales and the team replies within 24 hours.

Yes. On your first top-up, you get both the first-deposit bonus and the regular tier bonus for the amount you deposit, the two stack for maximum rewards.

No. All credits are valid forever, with no expiration. Whether you top up through self-serve or an enterprise plan, your balance never expires. Use the API at your own pace, you never have to worry about losing unused credit.

Self-serve tiers accept credit cards (Visa, Mastercard, JCB) and cryptocurrency (USDT / USDC). Business plans are prepaid and accept card, crypto, or bank wire transfer. Enterprise plans primarily use bank wire transfer, come with an invoice, and offer flexible payment terms.

Monthly usage of $20,000 USD or more, or any single top-up of $30,000 USD or more, qualifies for enterprise onboarding. Click Contact Sales and our team will reach out within 24 hours with a tailored per-model quote (up to 25% off select models) and a contract.

Rewards and commissions are platform API credit. Once transferred to your account balance they can be spent, and they are eligible for the daily spend-back program (see the pricing page for details). Cash withdrawal is not supported at this time.

Almost all models count toward commissions; a few specific models are excluded, and spending on them is not counted toward volume or commissions. See the reference list for details.

No. Share rates are calculated in tiers. A new rate applies only to volume earned after you cross the threshold; earlier volume is never recalculated.

Example: once your total volume reaches $8,000 and you hit Platinum, only volume after that earns 3%. So when your total volume reaches $8,200, only the $200 above the threshold earns 3%; everything up to $8,000 stays at the rate of the tier you were in at the time.

Once your friend completes their first top-up and spends $20, the invite reward is credited automatically. As a partner, last month’s commission is settled automatically on the 1st of each month.

Not at all. You don’t need to code or build a website. Just copy your unique link and share it on social, your blog, YouTube, or directly with friends to start promoting.

MixRoute is built and operated by ELITE CLOUD. Since 2023, ELITE CLOUD has served 2,100+ enterprise customers worldwide and holds 150+ cloud certifications as an official partner of AWS, GCP, and Azure.

Our business model works the same way the cloud reseller industry has for decades: we earn through supplier partnerships, not by marking up your bill. Billing follows official pricing, and usage is auditable line by line.

No. Your content exists in memory only while it is being processed. Nothing is written, logged, or used for training. We retain only the usage data required for billing and never touch your actual content.

Aggregation platforms typically add a 5–10% fee and route requests through the public queue. MixRoute charges 0% markup at official pricing, routes requests with higher rate limits instead of through the public queue, and fails over in milliseconds when a provider goes down so your service stays up.

Go to console.mixroute.ai to register and get one API key for 250+ mainstream models. It is OpenAI SDK compatible, so switching models takes just one parameter change.

Yes. Triage adds one extra step, and the time varies with request length. It’s a trade: a little time for 20–40% off your costs. For latency-critical endpoints, use a regular key to call a specific model directly and skip the router.

It reads, it never stores. Request content is classified in memory in real time, cleared immediately afterward, and passed through to the target model untouched. Nothing is written to persistent storage, consistent with MixRoute’s zero data retention policy.

Your model pool is the insurance: the router only ever chooses among models you’ve authorized and never assigns anything outside the pool. If some traffic must always hit one specific model, create a separate regular key for it and call that model directly.

One word: control. Most auto-routers are a total black box where the platform decides everything. MixRoute’s Smart Routing runs inside a model pool you define, with zero data retention and a savings dashboard. You always know where the money is saved and where your data goes.

No. Create a Smart Routing API key in the console and swap it in for your current key. The API call format stays exactly the same.

No. Smart Routing is available to everyone free of charge, with no routing service fee. You only pay for actual usage of the selected model, deducted from your credit balance and shown clearly on your bill.

We start from list price for each vendor you pick. Your rate follows your total monthly spend across all vendors. Volume pricing starts at $3,000 a month, with a better rate from $7,000. Above $10,000 a month, pricing is custom: contact us for bigger discounts, and we’ll email you a quote. For the comparison, we add OpenRouter’s 5.5% card top-up fee (it’s 5.0% on crypto and 8% on their Business plan). Your formal quote is final.

Most models run through Azure OpenAI, Google Vertex AI and AWS Bedrock. Same model, same version, never a cheaper substitute. Verify with your own prompts.

The base URL, the API key and the model IDs. The request format stays the same, since both APIs are OpenAI-compatible. Find each ID on our models page.

No. MixRoute credits never expire. On OpenRouter, unused credits can expire 365 days after purchase, per section 4.2 of their terms.

Free models right away, no card needed. Top up to unlock all 250+ models, and the credit works on all of them too.

No. Your prompts and outputs are not stored and not used for training. We keep usage data, such as token counts and timestamps, for billing. Each upstream provider applies its own data policy.

Then we’ll tell you not to switch. Send us your bill and we’ll compare it line by line, per model.

Yes. We set custom, per-model discounts on video and image generation models such as Seedance, Wan, Kling, MiniMax, Gemini and GPT-Image, based on which models you run and your monthly volume. Send us your usage and we’ll price each model within 24 hours. Get my custom discount

Per generated second, with the rate set by the model, the resolution and whether the clip has sound. Seedance is billed per video token, which also scales with resolution and length. Each model’s current rates are on its page at mixroute.ai/models.

Video generation is asynchronous. You submit a task to POST /v1/video/generations and get a task_id back right away, then poll GET /v1/video/generations/{task_id} every 10 to 15 seconds until it succeeds and returns the video URL. Kling uses its own native endpoints under /kling/. Each submission is a separate paid task, so don’t resubmit automatically after a timeout, and save results promptly because result links are temporary.

Video: Seedance 2.5 and 2.0 (plus 2.0 Fast and Mini); Wan 3.0 and 3.0 Prime, plus 2.7 and 2.6; Kling 3.0 and 3.0 Omni, plus 3.0 Turbo, O1, 2.6 and 2.5 Turbo; MiniMax H3; Gemini Omni Flash. Image: Nano Banana Pro, Nano Banana 2 and 2.1; Qwen-Image 2.0 and 2.0 Pro; GPT-Image-2.5 and GPT-Image-2; Grok Imagine 2.0. Exact IDs are on our models page.

Yes. Seedance 2.5 and 2.0, Wan 3.0 and 3.0 Prime, Kling 3.0 Omni, MiniMax H3 and Gemini Omni Flash all accept video as an input, for references, editing or extending a clip. Limits differ by model, for example Seedance 2.5 takes up to 10 reference videos and 30 seconds in total. Each model’s limits are in the video API docs.

No. Each clip and image was published by the team that built the model, and links to its source. Videos are compressed for the web. MixRoute serves the same models by ID through one API key, so you can run your own prompts and compare.

Yes. Any team can request 6,000 RPM, usually the same day, and Enterprise customers can go up to 300K RPM and 300M TPM. Tell us your expected peak and the models you’ll call, and we’ll confirm what we can provide.

Free models right away, no card needed. Top up to unlock all 250+ models, and the credit works on all of them too.

No. Your prompts and outputs are not stored and not used for training. We keep usage data, such as token counts and timestamps, for billing. Each upstream provider applies its own data policy.

Yes. Limits scale in both requests per minute and tokens per minute: 6,000 RPM on request, up to 300K RPM and 300M TPM per Enterprise customer. Tell us your peak and the models you use, and we’ll scope both.

Open a ticket in the console. Most upgrades to 6,000 RPM happen the same day. Higher rate limits beyond that are scoped with our team, typically within a week.

New accounts start at 50 RPM, rising to 100 RPM after the first top-up. Any team can request a higher rate limit.

That moves to Enterprise rate limits, up to 300K RPM and 300M TPM per customer. It’s dedicated throughput committed with upstream providers, scoped to your forecast and adjustable quarterly, evaluated case by case with our technical team.

Most models run through Azure OpenAI, Google Vertex AI and AWS Bedrock. Same model, same version, never a cheaper substitute. Verify with your own prompts.

No. Your prompts and outputs are not stored and not used for training. We keep usage data, such as token counts and timestamps, for billing. Each upstream provider applies its own data policy.

Standard plans are supported through tickets and your account manager, without an SLA. An SLA is available under an enterprise agreement.