Modelflare

Build beyond one model. Keep your product moving.

Modelflare is an OpenAI-compatible AI API gateway. One API key connects products and agents to GPT, Claude, Gemini, and leading image and video models, with provider routing, ordered fallback, and per-request usage and cost.

  • One OpenAI-compatible layer for LLM, image, and video models
  • Provider routing with ordered fallback routes
  • Request-level usage and cost visibility

From model access to reliable delivery.

  • More models. One place to connect.

    Compare models, capabilities, and prices across providers. Choose what fits each part of your product.

  • A backup route, built into your requests.

    When an upstream failure allows failover, try another eligible route for the same model. Configure a primary route and ordered backups for a regular API key.

  • One key. Your routing preference.

    Use the same key across models. Modelflare selects a route for the requested model from those available to your account, following your chosen strategy.

Quick start

API base URLs

  • OpenAI: https://modelflare.dev/v1
  • Anthropic: https://modelflare.dev
  • Gemini: https://modelflare.dev
curl https://modelflare.dev/v1/chat/completions \
  -H "Authorization: Bearer $MODELFLARE_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"model": "MODEL_ID", "messages": [{"role": "user", "content": "Hello"}]}'

Read the docs · View pricing

Enterprise & teams

Approved enterprise accounts get a console of their own: invite your team, give each member a budget, and see every top-up, request and creation in one place.

  • Members and budgets

    Invite admins and members by email, and give each an unlimited, one-time or monthly budget.

  • Keys with your prefix

    Enterprise keys start with your own short name, such as acme-, and every request is attributed to the member who made it.

  • Top-ups and alerts

    The owner and admins can top up. Alerts email you before the balance or a budget runs out.

  • Usage, creations and audit

    See spending by member, model and key, keep the images and videos members create, and review every admin action.

Explore enterprise plans

Common questions

What can I do with Modelflare?

Build with text, image, video, and audio models from different providers. One API and one console keep your keys, usage, and billing together.

Explore models

How is Modelflare different from calling each provider directly?

You use one API key, one base URL, and one balance for models from many providers. Each model lists its price next to the official rate, and every request records its token usage and cost.

Compare models

Is it easy to connect my existing app?

Usually you only change the base URL and API key. Use https://modelflare.dev/v1 with OpenAI SDKs and compatible tools, or https://modelflare.dev with Anthropic and Gemini SDKs. The docs cover each model and coding agents such as Claude Code and Codex.

Read the docs

How do I find the right model?

Start with what you want to make. On Models & Pricing, compare capabilities, supported APIs, and rates, then open a model for the details.

Compare models

How are prices calculated?

Each model lists its price for every routing group, in units such as 1M tokens, one image, or one second of video, beside the official rate. Your console records the usage and cost of each request.

View pricing

Do I need a subscription to start?

You can start with pay-as-you-go credits. If you prefer a plan, compare its included quota, supported models, and limits first. Your console shows usage and charges as you go.

View pricing

What happens when an upstream provider fails?

When an upstream failure allows failover, try another eligible route for the same model. Configure a primary route and ordered backups for a regular API key.

Read the docs

Can I keep each project under control?

Give each project its own API key. Set a quota and expiration date, add model or IP restrictions when needed, and track usage separately.

Manage API keys

Can my team share one account?

Yes. With an enterprise account, the whole team spends one shared balance. Invite members by email, give each a budget and their own keys, and see each member’s usage, requests and creations in the enterprise console. The owner and admins can top up, and alerts tell you before the balance or a budget runs out.

Explore enterprise plans

How does Modelflare handle my request data?

The Trust Center explains which request data is processed, how long records are kept, which service providers are involved, and how to report a security issue.

Trust Center

Can I create without writing code?

Of course. In Castaly, choose an image or video model and describe your idea. Sign in with your Modelflare account; your balance and usage records are shared.

Open Castaly