# KeepRouter

> KeepRouter is a managed multi-model API gateway with published customer pricing, prepaid pay-as-you-go credit, scoped API keys, compatible OpenAI and Anthropic chat routes, and route-specific endpoints for specialized modalities.

Use the live model catalog for current callable model IDs, endpoints, limits, makers, modalities, prices, and trial eligibility. Static pages deliberately do not preserve a model count or price snapshot.

## API entry points

- OpenAI-compatible base URL: `https://keeprouter.com/v1`
- Anthropic-compatible base URL: `https://keeprouter.com`
- Public OpenAPI: https://keeprouter.com/api/openapi.json
- Live model JSON: https://keeprouter.com/api/public/models
- Exact model JSON: https://keeprouter.com/api/public/model/gpt-6.1-sol (substitute a catalog ID; reviewed models include api_contract endpoint, reasoning and price-policy guidance)
- Interactive API reference: https://keeprouter.com/api/docs
- Quickstart: https://keeprouter.com/docs/quickstart.md

Compatibility is operation- and model-specific. Chat Completions, Responses, Messages, embeddings, images, audio, and other modalities are separate route contracts. Check the model page and test the exact production request instead of assuming universal feature parity.

## Live facts and machine-readable discovery

- [Models and current prices](https://keeprouter.com/models.md): live catalog with one Markdown page per canonical model ID.
- [Full reference](https://keeprouter.com/llms-full.txt): expanded product facts, live catalog, and the complete public content library.
- [AI index](https://keeprouter.com/ai-index.json): compact structured product and discovery manifest.
- [KeepRouter Skill](https://keeprouter.com/SKILL.md): agent instructions for authentication, model selection, bounded generation, and errors. Save as `keeprouter/SKILL.md` in your client's skill directory.
- [Sitemap](https://keeprouter.com/sitemap.xml): canonical public HTML URLs with freshness dates.
- [RSS feed](https://keeprouter.com/feed.xml): current KeepRouter editorial guides.
- Every product, audience, comparison, answer, blog, and selected documentation page below is available as Markdown at the same path plus `.md`, or by sending `Accept: text/markdown` to its HTML URL.

MCP and A2A are currently unavailable. Use the KeepRouter Skill with the documented REST API.

## Product features

- [Feature library](https://keeprouter.com/features.md)
- [Unified LLM API](https://keeprouter.com/features/unified-llm-api.md)
- [OpenAI-compatible API](https://keeprouter.com/features/openai-compatible-api.md)
- [Anthropic-compatible API](https://keeprouter.com/features/anthropic-compatible-api.md)
- [Explicit model routing](https://keeprouter.com/features/model-routing.md)
- [Usage-based billing](https://keeprouter.com/features/usage-based-billing.md)
- [API observability](https://keeprouter.com/features/api-observability.md)
- [Multimodal model routes](https://keeprouter.com/features/multimodal-models.md)

## Built for

- [Audience library](https://keeprouter.com/built-for.md)
- [AI startups](https://keeprouter.com/built-for/ai-startups.md)
- [Agent builders](https://keeprouter.com/built-for/agent-builders.md)
- [Developer tools](https://keeprouter.com/built-for/devtools.md)
- [Engineering teams](https://keeprouter.com/built-for/engineering-teams.md)
- [AI prototyping](https://keeprouter.com/built-for/ai-prototyping.md)
- [Retrieval and RAG applications](https://keeprouter.com/built-for/retrieval-applications.md)
- [Customer support agents](https://keeprouter.com/built-for/customer-support-agents.md)

## Comparisons

- [Comparison library](https://keeprouter.com/compare.md)
- [Best AI gateways: a fit-based buyer guide](https://keeprouter.com/compare/best-ai-gateways.md)
- [OpenRouter alternatives](https://keeprouter.com/compare/openrouter-alternatives.md)
- [Managed vs self-hosted AI gateways](https://keeprouter.com/compare/managed-vs-self-hosted-ai-gateways.md)
- [AI gateway vs direct provider APIs](https://keeprouter.com/compare/direct-provider-apis.md)
- [KeepRouter vs OpenRouter](https://keeprouter.com/compare/openrouter.md)
- [KeepRouter vs Portkey](https://keeprouter.com/compare/portkey.md)
- [KeepRouter vs LiteLLM](https://keeprouter.com/compare/litellm.md)
- [KeepRouter and Cloudflare AI Gateway](https://keeprouter.com/compare/cloudflare-ai-gateway.md)
- [KeepRouter vs Together AI](https://keeprouter.com/compare/together-ai.md)
- [KeepRouter vs Fireworks AI](https://keeprouter.com/compare/fireworks-ai.md)
- [KeepRouter and Vercel AI Gateway](https://keeprouter.com/compare/vercel-ai-gateway.md)
- [KeepRouter and Helicone](https://keeprouter.com/compare/helicone.md)
- [KeepRouter and Kong AI Gateway](https://keeprouter.com/compare/kong-ai-gateway.md)
- [KeepRouter and Amazon Bedrock](https://keeprouter.com/compare/amazon-bedrock.md)
- [KeepRouter and Gemini Enterprise Agent Platform](https://keeprouter.com/compare/gemini-enterprise-agent-platform.md)
- [KeepRouter and Microsoft Foundry](https://keeprouter.com/compare/microsoft-foundry.md)
- [KeepRouter and Hugging Face Inference Providers](https://keeprouter.com/compare/hugging-face-inference-providers.md)
- [Vercel AI Gateway vs OpenRouter](https://keeprouter.com/compare/vercel-ai-gateway-vs-openrouter.md)

Comparisons use current official documentation and explain when either operating model is the better fit. They do not claim universal winners, feature parity, rankings, or benchmark results.

## Direct answers

- [Answer library](https://keeprouter.com/answers.md)
- [What is an AI gateway?](https://keeprouter.com/answers/what-is-an-ai-gateway.md)
- [What is an OpenAI-compatible API?](https://keeprouter.com/answers/what-is-an-openai-compatible-api.md)
- [How do I use one API for multiple LLMs?](https://keeprouter.com/answers/how-to-use-one-api-for-multiple-llms.md)
- [How can I switch LLM models without changing code?](https://keeprouter.com/answers/how-to-switch-llm-models-without-changing-code.md)
- [How does LLM API pricing work?](https://keeprouter.com/answers/how-does-llm-api-pricing-work.md)
- [Can Claude Code use non-Claude models?](https://keeprouter.com/answers/can-claude-code-use-non-claude-models.md)
- [What is LLM routing?](https://keeprouter.com/answers/what-is-llm-routing.md)
- [AI gateway vs API gateway](https://keeprouter.com/answers/ai-gateway-vs-api-gateway.md)
- [Does an AI gateway add latency?](https://keeprouter.com/answers/does-ai-gateway-add-latency.md)
- [Is an AI gateway secure?](https://keeprouter.com/answers/is-an-ai-gateway-secure.md)
- [What is LLM API failover?](https://keeprouter.com/answers/what-is-llm-api-failover.md)
- [How to choose an AI gateway](https://keeprouter.com/answers/how-to-choose-an-ai-gateway.md)
- [What is a BYOK AI gateway?](https://keeprouter.com/answers/what-is-byok-ai-gateway.md)
- [When should you use an AI gateway?](https://keeprouter.com/answers/when-should-you-use-an-ai-gateway.md)
- [How to use OpenRouter with Python](https://keeprouter.com/answers/how-to-use-openrouter-with-python.md)
- [Does OpenRouter store prompts?](https://keeprouter.com/answers/does-openrouter-store-prompts.md)

## Guides and editorial research

- [Blog index](https://keeprouter.com/blog.md)
- [AI gateway guide](https://keeprouter.com/blog/ai-gateway-guide.md)
- [OpenAI-compatible API migration checklist](https://keeprouter.com/blog/openai-compatible-api-migration-checklist.md)
- [LLM routing versus load balancing](https://keeprouter.com/blog/llm-routing-vs-load-balancing.md)
- [Control multi-model API costs](https://keeprouter.com/blog/control-multi-model-api-costs.md)
- [Responses API versus Chat Completions](https://keeprouter.com/blog/responses-api-vs-chat-completions.md)
- [How to evaluate an AI gateway](https://keeprouter.com/blog/evaluate-ai-gateway.md)
- [Migrating from OpenRouter](https://keeprouter.com/blog/openrouter-alternatives-for-developers.md)
- [Kimi K3 long-context guide](https://keeprouter.com/blog/kimi-k3-1m-context.md)
- [Measure Claude Code model costs](https://keeprouter.com/blog/cut-claude-code-costs.md)
- [AI gateway security checklist](https://keeprouter.com/blog/ai-gateway-security-checklist.md)
- [AI gateway latency measurement guide](https://keeprouter.com/blog/ai-gateway-latency-guide.md)
- [LLM failover design guide](https://keeprouter.com/blog/llm-failover-design-guide.md)
- [Async video generation API](https://keeprouter.com/blog/async-video-generation-api.md)
- [Choosing a video model tier](https://keeprouter.com/blog/choosing-a-video-model-tier.md)
- [Text-to-speech billing per second](https://keeprouter.com/blog/text-to-speech-per-second-billing.md)
- [Multimodal embeddings guide](https://keeprouter.com/blog/multimodal-embeddings-guide.md)
- [Webhooks for long-running AI jobs](https://keeprouter.com/blog/long-running-job-webhooks.md)
- [Image-to-video and video editing APIs](https://keeprouter.com/blog/image-to-video-api-guide.md)
- [Auditing cached input pricing](https://keeprouter.com/blog/llm-cache-pricing-audit.md)
- [Embedding model migration and reindex checklist](https://keeprouter.com/blog/embedding-model-migration-reindex-checklist.md)
- [Support-agent cost per verified resolution](https://keeprouter.com/blog/support-agent-cost-per-resolution.md)

## 简体中文内容

The five editorial families have stable Simplified Chinese HTML and Markdown routes. Chinese pages use a `/zh` prefix, self-canonical URLs, `hreflang="zh-CN"`, localized structured data, and language-preserving internal links.

- [中文功能库](https://keeprouter.com/zh/features.md)
- [中文适用团队内容库](https://keeprouter.com/zh/built-for.md)
- [中文产品对比与选型库](https://keeprouter.com/zh/compare.md)
- [中文问答库](https://keeprouter.com/zh/answers.md)
- [中文博客](https://keeprouter.com/zh/blog.md)
- [包含全部中英文页面的完整参考](https://keeprouter.com/llms-full.txt)

## Implementation and trust

- [Claude Code setup](https://keeprouter.com/use-cases/claude-code.md)
- [OpenAI SDK setup](https://keeprouter.com/use-cases/openai-sdk.md)
- [Continue.dev config.yaml setup](https://keeprouter.com/use-cases/continue-dev.md)
- [API errors](https://keeprouter.com/docs/errors)
- [Security and data handling](https://keeprouter.com/security)
- [Service status](https://keeprouter.com/status)
- [Changelog](https://keeprouter.com/changelog)
- [LLM API glossary](https://keeprouter.com/glossary.md)
- [About KeepRouter](https://keeprouter.com/about.md)

The service does not write prompt or completion bodies to request logs. Optional response caching is off by default. The public status page reports KeepRouter database, schema, and routing-configuration health; it does not claim a fresh upstream inference probe for every model.

## Getting started

1. Sign in with an email verification code.
2. Create a key restricted to the currently published `free` model.
3. Run the quickstart without a top-up and inspect the measured request.
4. Choose a paid model from the live catalog, set an appropriate model allowlist and spending boundary, and add prepaid USD credit before calling it.

## Contact

- support@keeprouter.com

## Cost planning
- [API cost calculator](https://keeprouter.com/tools/api-cost-calculator.md): Calculate model usage from current customer prices, with separate token, image, video and audio units.
- [API migration checker](https://keeprouter.com/tools/api-migration-checker.md): Check credential-free JSON locally against current model IDs and endpoint contracts. Flags OpenRouter routing differences and tests still required; no inference calls.

## Model pricing and migration guides

- [DeepSeek API pricing: estimate cache hits without counting input twice](https://keeprouter.com/blog/deepseek-api-pricing-cache-estimation.md)
- [DeepSeek API 价格：缓存命中后，怎样不重复计算输入费用](https://keeprouter.com/zh/blog/deepseek-api-pricing-cache-estimation.md)
- [Claude API cost: plan prompt caching around writes, reads and reuse](https://keeprouter.com/blog/claude-api-cost-prompt-caching.md)
- [Claude API 费用：按写入、读取和复用次数规划 Prompt Caching](https://keeprouter.com/zh/blog/claude-api-cost-prompt-caching.md)
- [GPT API pricing and SDK migration: keep the request contract visible](https://keeprouter.com/blog/gpt-api-pricing-sdk-migration.md)
- [GPT API 价格与 SDK 迁移：把请求契约和费用一起核对](https://keeprouter.com/zh/blog/gpt-api-pricing-sdk-migration.md)
- [Gemini with the OpenAI SDK: compare the native, compatible and gateway paths](https://keeprouter.com/blog/gemini-openai-compatible-api-differences.md)
- [用 OpenAI SDK 调用 Gemini：区分原生、兼容端点和网关路径](https://keeprouter.com/zh/blog/gemini-openai-compatible-api-differences.md)
- [Doubao multimodal embeddings: build a small image-and-text retrieval test](https://keeprouter.com/blog/doubao-multimodal-embedding-retrieval.md)
- [Doubao 多模态 Embedding：建立一个图文检索小样本](https://keeprouter.com/zh/blog/doubao-multimodal-embedding-retrieval.md)
- [OpenRouter to KeepRouter migration: map URLs, model IDs and routing fields](https://keeprouter.com/blog/openrouter-to-keeprouter-migration.md)
- [从 OpenRouter 迁移到 KeepRouter：核对 URL、模型 ID 与路由字段](https://keeprouter.com/zh/blog/openrouter-to-keeprouter-migration.md)
- [KeepRouter API key setup: from the free model to a controlled paid request](https://keeprouter.com/blog/keeprouter-api-key-free-to-paid.md)
- [KeepRouter API Key 配置：从免费模型走到一次受控付费请求](https://keeprouter.com/zh/blog/keeprouter-api-key-free-to-paid.md)
- [OpenRouter vs LiteLLM: compare operating cost for your actual workload](https://keeprouter.com/blog/openrouter-vs-litellm-operating-cost.md)
- [OpenRouter 与 LiteLLM：按实际工作负载比较总运行成本](https://keeprouter.com/zh/blog/openrouter-vs-litellm-operating-cost.md)

- [Gemini API: model selection, OpenAI SDK and pricing](https://keeprouter.com/blog/gemini-api-models-pricing-guide.md)
- [Gemini API：型号选择、OpenAI SDK 与价格](https://keeprouter.com/zh/blog/gemini-api-models-pricing-guide.md)

## Application integration and API operations guides

- [n8n OpenAI-Compatible API: Connect and Debug a Workflow](https://keeprouter.com/blog/n8n-openai-compatible-api.md)
- [n8n 接入 OpenAI 兼容 API：配置与工作流排障](https://keeprouter.com/zh/blog/n8n-openai-compatible-api.md)
- [Dify Custom Model Setup with an OpenAI-Compatible API](https://keeprouter.com/blog/dify-openai-compatible-model.md)
- [Dify 自定义模型：接入 OpenAI 兼容 API](https://keeprouter.com/zh/blog/dify-openai-compatible-model.md)
- [LangChain Custom Base URL: An OpenAI-Compatible API Guide](https://keeprouter.com/blog/langchain-openai-compatible-api.md)
- [LangChain 自定义 Base URL：兼容 API 接入指南](https://keeprouter.com/zh/blog/langchain-openai-compatible-api.md)
- [LlamaIndex OpenAILike: Change the LLM, Keep the RAG Index](https://keeprouter.com/blog/llamaindex-openai-like-rag.md)
- [LlamaIndex OpenAILike：更换 LLM，保留 RAG 索引](https://keeprouter.com/zh/blog/llamaindex-openai-like-rag.md)
- [Cline OpenAI-Compatible Setup: Models, Tools and Cost](https://keeprouter.com/blog/cline-openai-compatible-api.md)
- [Cline 接入 OpenAI 兼容 API：模型、工具与费用](https://keeprouter.com/zh/blog/cline-openai-compatible-api.md)
- [Roo Code Migration: Keep Your API, Replace the Client](https://keeprouter.com/blog/roo-code-migration.md)
- [Roo Code 停运后迁移：保留 API，替换代码客户端](https://keeprouter.com/zh/blog/roo-code-migration.md)
- [Continue API Setup: Chat and Autocomplete Roles](https://keeprouter.com/blog/continue-chat-autocomplete-models.md)
- [Continue API 配置：分清聊天与自动补全角色](https://keeprouter.com/zh/blog/continue-chat-autocomplete-models.md)
- [Aider with an OpenAI-Compatible API: Setup and Edit Tests](https://keeprouter.com/blog/aider-openai-compatible-api.md)
- [Aider 接入 OpenAI 兼容 API：配置与编辑验证](https://keeprouter.com/zh/blog/aider-openai-compatible-api.md)
- [Open WebUI Custom API: Connect Models Without Surprises](https://keeprouter.com/blog/open-webui-openai-compatible-api.md)
- [Open WebUI 自定义 API：连接模型与排查功能差异](https://keeprouter.com/zh/blog/open-webui-openai-compatible-api.md)
- [LibreChat Custom Endpoint: Models, Titles and API Costs](https://keeprouter.com/blog/librechat-custom-endpoint.md)
- [LibreChat 自定义端点：模型、标题与 API 费用](https://keeprouter.com/zh/blog/librechat-custom-endpoint.md)
- [Structured Outputs vs JSON Mode: Validate Real Data](https://keeprouter.com/blog/structured-outputs-json-mode.md)
- [Structured Outputs 与 JSON Mode：如何验证真实数据](https://keeprouter.com/zh/blog/structured-outputs-json-mode.md)
- [LLM Tool Calling: Build the Full Request-Result Loop](https://keeprouter.com/blog/llm-tool-calling-loop.md)
- [LLM 工具调用：实现完整的请求与结果循环](https://keeprouter.com/zh/blog/llm-tool-calling-loop.md)
- [OpenAI-Compatible API 429 Errors: Limits, Quota and Retry](https://keeprouter.com/blog/openai-compatible-api-429-errors.md)
- [OpenAI 兼容 API 的 429 错误：限流、额度与重试](https://keeprouter.com/zh/blog/openai-compatible-api-429-errors.md)
- [LLM Streaming Stops Early: Diagnose SSE and Completion](https://keeprouter.com/blog/llm-streaming-interrupted.md)
- [LLM 流式输出中断：区分 SSE、超时和完成状态](https://keeprouter.com/zh/blog/llm-streaming-interrupted.md)
- [Context Window Exceeded: Budget Prompts Before Retrying](https://keeprouter.com/blog/context-window-exceeded-fix.md)
- [上下文超限怎么办：先核算提示词，再决定重试](https://keeprouter.com/zh/blog/context-window-exceeded-fix.md)
- [RAG Cost per Query: A Worksheet Beyond Token Prices](https://keeprouter.com/blog/rag-api-cost-per-query.md)
- [RAG 每次查询多少钱：别只计算生成 token](https://keeprouter.com/zh/blog/rag-api-cost-per-query.md)
- [Reasoning Token Costs: Read Usage Without Double Counting](https://keeprouter.com/blog/reasoning-tokens-api-cost.md)
- [推理 token 怎么计费：避免重复计算与短回答误判](https://keeprouter.com/zh/blog/reasoning-tokens-api-cost.md)
- [GPT vs Claude vs Gemini for Document Extraction: Test Plan](https://keeprouter.com/blog/gpt-claude-gemini-document-extraction.md)
- [GPT、Claude、Gemini 文档提取选型：样本与评分方法](https://keeprouter.com/zh/blog/gpt-claude-gemini-document-extraction.md)
- [Batch API vs Real-Time: When the Discount Is Worth It](https://keeprouter.com/blog/batch-api-vs-realtime-cost.md)
- [Batch API 与实时请求：什么时候批处理更划算](https://keeprouter.com/zh/blog/batch-api-vs-realtime-cost.md)
- [Ollama vs Hosted APIs: Cost, Capacity and Break-Even](https://keeprouter.com/blog/ollama-vs-hosted-api-cost.md)
- [Ollama 本地模型与托管 API：成本、容量与盈亏平衡](https://keeprouter.com/zh/blog/ollama-vs-hosted-api-cost.md)

## October 2026 model integration guides

- [GPT-6.1 Sol API: Responses, tools and a reliable migration](https://keeprouter.com/blog/gpt-6-1-sol-api-tools-pricing.md)
- [GPT-6.1 Sol API：Responses、工具调用与可靠迁移](https://keeprouter.com/zh/blog/gpt-6-1-sol-api-tools-pricing.md)
- [Claude Sonnet 5.5 and Opus 5.5: API migration and cache costs](https://keeprouter.com/blog/claude-sonnet-opus-5-5-api-migration.md)
- [Claude Sonnet 5.5 与 Opus 5.5：API 迁移与缓存成本](https://keeprouter.com/zh/blog/claude-sonnet-opus-5-5-api-migration.md)
- [MiMo V2.6 API: Flash, Pro, UltraSpeed and the V2.5 migration](https://keeprouter.com/blog/mimo-v2-6-api-pricing-migration.md)
- [MiMo V2.6 API：Flash、Pro、UltraSpeed 与 V2.5 迁移](https://keeprouter.com/zh/blog/mimo-v2-6-api-pricing-migration.md)
- [GLM 5.3 vs Flash vs FlashX: inputs, reasoning and API costs](https://keeprouter.com/blog/glm-5-3-flash-flashx-api-guide.md)
- [GLM 5.3、Flash 与 FlashX：输入、思考与 API 成本](https://keeprouter.com/zh/blog/glm-5-3-flash-flashx-api-guide.md)
- [Grok 4.7, Qwen 3.8, Gemma 4 and Doubao 2.1: choose by task](https://keeprouter.com/blog/grok-qwen-gemma-doubao-model-selection.md)
- [Grok 4.7、Qwen 3.8、Gemma 4 与 Doubao 2.1：按任务选型](https://keeprouter.com/zh/blog/grok-qwen-gemma-doubao-model-selection.md)