# flatkey.ai

flatkey.ai is a unified AI API gateway, model routing, billing, and operations platform.

## API Protocols

- Router base URL: https://router.flatkey.ai
- OpenAI-compatible Chat Completions: POST https://router.flatkey.ai/v1/chat/completions
- OpenAI-compatible Responses: POST https://router.flatkey.ai/v1/responses
- Anthropic Messages: POST https://router.flatkey.ai/v1/messages
- Gemini native generateContent: POST https://router.flatkey.ai/v1beta/models/{model}:generateContent
- Asynchronous video generation: POST https://router.flatkey.ai/v1/videos

## Gemini Image Models

- nano-banana-pro-preview supports both Gemini native generateContent and OpenAI-compatible Chat Completions.
- With Gemini native generateContent, generated images are returned in candidates[].content.parts[].inlineData.
- With Chat Completions, generated images are returned in choices[0].message.content as Markdown data URIs.
- These Gemini image-model paths currently do not use /v1/images/generations; do not infer lack of Gemini support from that endpoint.

## Seedance Video Models

- Seedance models use the official multimodal content[] request format.
- Create a video task with POST https://router.flatkey.ai/v1/videos.
- Poll task status with GET https://router.flatkey.ai/v1/videos/{task_id}.
- Download completed video content with GET https://router.flatkey.ai/v1/videos/{task_id}/content.
- Use GET /v1/models to discover the Seedance model IDs currently available to your account.
- Documentation: https://docs.flatkey.ai/api-reference/seedance-video-generation
- Chinese documentation: https://docs.flatkey.ai/zh/api-reference/seedance-video-generation

## Core Pages

- Home: https://flatkey.ai/
- Flatkey Tools: https://flatkey.ai/tools
- Flatkey as an Apify alternative: https://flatkey.ai/apify-alternative
- Flatkey CLI: https://flatkey.ai/cli
- Flatkey vs Higgsfield: https://flatkey.ai/higgsfield-alternative
- Model pricing: https://flatkey.ai/pricing
- Rankings: https://flatkey.ai/rankings
- Blog: https://flatkey.ai/blog
- Sitemap: https://flatkey.ai/sitemap.xml

## Blog Categories

- AI Gateway Architecture: https://flatkey.ai/blog/category/ai-gateway-architecture
- Base URL and SDK Migration: https://flatkey.ai/blog/category/base-url-and-sdk-migration
- Cost, Billing, and Ops: https://flatkey.ai/blog/category/cost-billing-and-ops
- Enterprise Controls and Trust: https://flatkey.ai/blog/category/enterprise-controls-and-trust
- Gateway Comparisons: https://flatkey.ai/blog/category/gateway-comparisons
- Model and Modality Playbooks: https://flatkey.ai/blog/category/model-and-modality-playbooks
- Reliability and Routing: https://flatkey.ai/blog/category/reliability-and-routing
- Tool Integrations: https://flatkey.ai/blog/category/tool-integrations
- コスト・請求・運用: https://flatkey.ai/blog/category/cost-billing-ops-ja

## Blog Articles

- Model Fallback Strategy: A 3-Workflow Playbook: https://flatkey.ai/blog/model-fallback-strategy-three-workflow-playbook - A production model fallback playbook with policy as code, failure drills, a 60-minute game day, release gates, rollback triggers, and safe recovery rules.
- AI API Cost Optimization: 7 Strategies, 5 Alternatives, and a Cost Calculator: https://flatkey.ai/blog/ai-api-cost-optimization-strategies-alternatives - Calculate cost per accepted task, compare five AI API alternatives with a 100-task benchmark, and run a safer cost reduction sprint.
- LLM Gateway Beginner Guide: From First Request to Production: https://flatkey.ai/blog/llm-gateway-beginner-guide - A practical LLM gateway beginner guide with a quickstart, first-100-requests lab, error map, build-versus-buy scorecard, and production rollout checks.
- AI Observability Implementation Checklist: 20 Production Steps: https://flatkey.ai/blog/ai-observability-implementation-checklist - A production AI observability implementation checklist with 20 launch steps, a telemetry contract, code pattern, alert runbook, acceptance tests, and ownership handoff.
- Prompt Caching Workflow: Cost and ROI Guide for LLM Apps: https://flatkey.ai/blog/prompt-caching-workflow-cost-roi-guide - A provider-aware prompt caching workflow with ROI formulas, a seven-day audit, break-even math, telemetry, and rollout guardrails for production LLM applications.
- What Is an LLM Gateway? A Beginner’s Guide: https://flatkey.ai/blog/what-is-an-llm-gateway - Learn how an LLM gateway centralizes model access, routing, fallback, rate limits, observability, security, and billing for AI applications.
- LLM API Observability: Metrics, Traces, Logs, and Cost: https://flatkey.ai/blog/llm-api-observability-metrics-traces-cost - A production guide to monitoring LLM APIs with validated success metrics, distributed traces, safe structured logs, SLOs, alerts, and cost per accepted task.
- Secure API Key Management for AI Products: https://flatkey.ai/blog/secure-api-key-management-ai-products - A production playbook for AI API key custody, scoped identities, prompt-safe logging, zero-downtime rotation, leak response, and multi-provider governance.
- AI Model Evaluation Before Switching API Providers: A Workflow Checklist: https://flatkey.ai/blog/ai-model-evaluation-before-switching-api-providers - Use a decision-grade AI model evaluation workflow to compare providers on task success, compatibility, reliability, latency, effective cost, and rollout risk.
- LLM Rate Limits Explained: RPM, TPM, and Retries: https://flatkey.ai/blog/llm-rate-limits-explained-rpm-tpm-retries - Understand RPM, TPM, 429 errors, capacity planning, queues, exponential backoff, retry budgets, and fallback routing for production LLM APIs.
- LLM API Fallback Routing: A Production Failover Playbook: https://flatkey.ai/blog/llm-api-fallback-routing-production-playbook - A production playbook for deciding when LLM requests should retry, fail over, switch models, or stop—without breaking streams, tools, schemas, or latency budgets.
- AI API Pricing Comparison: OpenAI vs Claude vs Gemini vs Qwen (2026): https://flatkey.ai/blog/ai-api-pricing-comparison-openai-claude-gemini-qwen - Compare current OpenAI, Claude, Gemini, and Qwen API prices using normalized production workloads and cost per accepted result.
- AI API Gateway Architecture: One Key, Model Routing, and Failover: https://flatkey.ai/blog/ai-api-gateway-architecture-model-routing - A production architecture guide to one-key model access, explicit routing policy, health checks, retries, contract-safe failover, streaming, telemetry, and migration.
- Seedance API Evaluation Framework for Text-to-Video Product Teams: https://flatkey.ai/blog/seedance-api-evaluation-framework-product-teams - A repeatable framework for evaluating Seedance API quality, reliability, user experience, safety, and cost before a text-to-video product rollout.
- DeepSeek API vs Qwen API for Cost-Sensitive Workflows: 2026 Cost Guide: https://flatkey.ai/blog/deepseek-api-vs-qwen-api-cost-sensitive-workflows - Compare DeepSeek and Qwen API pricing with July 28, 2026 list prices, cache break-even math, batch costs, lifecycle risks, and workload-specific guidance.
- Gemini API Production Readiness Checklist for Backend Teams: https://flatkey.ai/blog/gemini-api-production-readiness-checklist-backend-teams - A production-readiness checklist for backend teams integrating Gemini API, covering credentials, response contracts, retries, observability, cost, rollout, and incidents.
- Claude API Access Outside One-Region Setups: A Compliance-First Guide: https://flatkey.ai/blog/claude-api-access-outside-one-region-setups - A compliance-first guide to Claude API access across regions, covering direct Anthropic, Bedrock, Vertex AI, gateways, testing, observability, and approved failover.
- OpenAI API Access for Multi-Model Products: A Production Setup Guide: https://flatkey.ai/blog/openai-api-access-production-multi-model-products - Set up OpenAI API access for a production multi-model product with project-scoped credentials, endpoint checks, rate-limit handling, canaries, and fallback readiness.