HTTP 200 verified twice
AI infrastructure · Tool
OpenRouter
ResearchedOpenRouter is a unified AI infrastructure platform providing access to 400+ models from 70+ providers through a single OpenAI-compatible API. It handles fallbacks, billing, and security for chat, image, embedding, and transcription workloads.
Official site snapshots
Visit live site See the official site at a glance
Read-only public captures of OpenRouter’s homepage. Screenshots are dated, never live embeds, and open full-screen.
1 earlier capture
In one minute
Start here for the decision-making essentials: what OpenRouter does, who it is for, how it is accessed, and the first-party sources behind this profile.
Pricing6 options
Credit-based pay-as-you-go model with no subscriptions
credits (e.g., $10 or $99 top-ups
Three plans offered: Free, Pay-as-you-go, and Enterprise
Pay-as-you-go tier carries a 5.5% platform fee, with fee discounts available
Small fee charged when purchasing credits
model inference pricing is passed through from
Platforms
Web
API accessNot public
FoundedStarted in early 2023
AvailabilityWeb / remote
LicenseProprietary
Best suited to
Source-backed fit Development teams building multi-model LLM applications Enterprises needing unified access across AI providers Startups wanting an OpenAI-compatible API without vendor lock-in Organizations requiring fine-grained data and security controls 10M+ global users and 250k+ apps using OpenRouter Millions of developers
Decision support
6 sourced answersCommon questions and adoption checks
Short answers to the questions buyers and builders commonly ask about OpenRouter. Each answer cites the shared ledger below, where every source is listed once.
01What does OpenRouter say it can do?
Unified interface providing access to many LLMs through a single API · Provides access to 400+ AI models · Provides a unified API giving access to hundreds of AI models through a single endpoint, w · Unified API to access major LLM models with aggregate billing and usage analytics
The Unified Interface For LLMs
02Who is OpenRouter intended for?
10M+ global users and 250k+ apps using OpenRouter · Millions of developers
10M+ Global Users
03What use cases does OpenRouter describe?
Illustration, anime, painting, and other expressive artistic styles · Coding, repository-level changes, long-horizon problem solving, and agentic workflows · Software engineering and agentic coding use cases · Rapid iteration and graphic design exploration where fast generation is the priority
particular strengths in illustration, anime, painting, and other expressive artistic styles
04What should teams verify before adopting OpenRouter?
Free tier rate limit is 50 requests per day
Rate limits 50 reqs/day
Ledger citation[2] openrouter.ai/pricing
05What pricing information is available for OpenRouter?
Credit-based pay-as-you-go model with no subscriptions; credits (e.g., $10 or $99 top-ups · Three plans offered: Free, Pay-as-you-go, and Enterprise · Pay-as-you-go tier carries a 5.5% platform fee, with fee discounts available · Small fee charged when purchasing credits; model inference pricing is passed through from
Credits can be used with any model or provider.
06Does OpenRouter document API access?
Single API that is fully OpenAI compatible, working with the OpenAI SDK out of the box · Standard HTTP endpoint at /api/v1/chat/completions, compatible with any language or framew · Client SDKs available for type-safe model calls with minimal overhead · Agent SDK for building agents with tool use, loops, and state
OpenAI SDK works out of the box.
Decision guide
AI infrastructureCapabilities and operating fit
This profile connects the jobs OpenRouter is described as handling with its delivery model, access options and the subjects used to match it to related products in this directory.
Common use cases
- Multi-model LLM orchestration with automatic provider fallbacks
- Chat, image generation, embeddings, and transcription via one endpoint
- Building agents with tool use, loops, and state management
- Routing traffic across providers for higher availability
- Accessing newest flagship models without code redeployment
- Enterprise deployments with custom data policies and Private Models
Topics mapped
Verified capabilities
- Unified interface providing access to many LLMs through a single API
- Provides access to 400+ AI models
- Provides a unified API giving access to hundreds of AI models through a single endpoint, w
- Unified API to access major LLM models with aggregate billing and usage analytics
- Pass-through provider pricing with pooled uptime and automatic fallbacks for reliability
- Routes 100 trillion tokens per month across the platform
- Speech-to-text is live with Whisper, GPT-4o Mini Transcribe, and Voxtral; text-to-speech e
- Model Fusion routes prompts to multiple models in parallel and synthesizes their responses
Recorded integrations
Access to 70+ providers including OpenAI, Anthropic, and MetaSupports integrations with Effect AI SDK, Arize, AX LangChain, LiveKit, Langfuse, Mastra, Simple integration via a single slug to support new models without separate provider integRoutes through 7+ AI providers with automatic fallback if one provider goes downIntegrates with 70+ providersOpenRouter offers an MCP Server that connects a coding agent to the live model catalog, be
Intended audiences
Access signals
- Pricing model
- Credit-based pay-as-you-go model with no subscriptions; credits (e.g., $10 or $99 top-ups · Three plans offered: Free, Pay-as-you-go, and Enterprise · Pay-as-you-go tier carries a 5.5% platform fee, with fee discounts available · Small fee charged when purchasing credits; model inference pricing is passed through from
- API
- Not publicly listed
- Source links
- 17 recorded
Source-backed
Updated July 22, 2026Verified facts
Each fact points to a recorded source, making it easy to distinguish verified product information from claims that need checking.
OpenRouter
Unified interface providing access to many LLMs through a single API · Access to 400+ AI models · 70+ providers including OpenAI, Anthropic, and Meta
Unified interface providing access to many LLMs through a single API
Access to 70+ providers including OpenAI, Anthropic, and Meta
Single API that is fully OpenAI compatible, working with the OpenAI SDK out of the box
Credit-based pay-as-you-go model with no subscriptions; credits (e.g., $10 or $99 top-ups shown) work with any model or provider
Supports chat, image generation, embeddings, and transcription through one base URL
View 32 more verified facts
Runs at the edge with fallback to other providers when one goes down for higher availability
Custom fine-grained data policies to restrict prompts to trusted models and providers
OpenRouter, Inc
10M+ global users and 250k+ apps using OpenRouter
Three plans offered: Free, Pay-as-you-go, and Enterprise
Pay-as-you-go tier carries a 5.5% platform fee, with fee discounts available
Provides access to 400+ AI models
Connects to 70+ providers (4 free providers and 25+ free models available)
Tiered support: Community Support (Free), Email Support (Pay-as-you-go), Email Support plus Support SLA with Shared Slack Channel (Enterprise)
Free tier rate limit is 50 requests per day
Provides a unified API giving access to hundreds of AI models through a single endpoint, with automatic fallbacks and cost-effective model selection
Standard HTTP endpoint at /api/v1/chat/completions, compatible with any language or framework
Client SDKs available for type-safe model calls with minimal overhead
Agent SDK for building agents with tool use, loops, and state
Supports integrations with Effect AI SDK, Arize, AX LangChain, LiveKit, Langfuse, Mastra, OpenAI SDK, Anthropic Agent SDK, PydanticAI, Replit, TanStack AI, Vercel AI SDK, Xcode, Zapier, and Infisical
Supports latest model aliases (e.g., ~openai/gpt-latest) that always resolve to the newest flagship models without redeployment
Unified API to access major LLM models with aggregate billing and usage analytics
Pass-through provider pricing with pooled uptime and automatic fallbacks for reliability
Small fee charged when purchasing credits; model inference pricing is passed through from providers without markup
Per-model pricing that may differ for prompt/completion, images, and reasoning tokens
BYOK usage includes a free monthly request allowance; subsequent usage incurs a small percentage fee based on equivalent OpenRouter cost
Community support via Discord #help forum
Closed a $113M Series B funding round
Routes 100 trillion tokens per month across the platform
Workspace Guardrails enforce zero data retention, block prompt injection against 30+ OWASP-derived patterns, and redact PII before it reaches a provider
Speech-to-text is live with Whisper, GPT-4o Mini Transcribe, and Voxtral; text-to-speech exposes supported_voices in the models API
Model Fusion routes prompts to multiple models in parallel and synthesizes their responses into a single answer, available as an API plugin, a server tool, and in the chatroom composer
Private Models (custom, fine-tuned, or dedicated endpoints) are available exclusively on the Enterprise plan
Simple integration via a single slug to support new models without separate provider integrations
Routes through 7+ AI providers with automatic fallback if one provider goes down
Provides access to 28+ AI models, letting users choose by quality, speed, or price
Stealth models give users access to the newest models the moment they drop, with minimal developer burden
Practical capabilities
8 documented areasWhat it helps with
A concise view of the jobs, capabilities and integrations described in the recorded product sources.
Multi-model LLM orchestration with automatic provider fallbacks
Chat, image generation, embeddings, and transcription via one endpoint
Building agents with tool use, loops, and state management
Routing traffic across providers for higher availability
Accessing newest flagship models without code redeployment
Enterprise deployments with custom data policies and Private Models
Illustration, anime, painting, and other expressive artistic styles
Coding, repository-level changes, long-horizon problem solving, and agentic workflows
Availability
Source checkedWhere it runs and where to get it
Documented product formats, platforms and official distribution destinations. Availability can vary by region and plan.
Cost / license
Credit-based pay-as-you-go model with no subscriptions; credits (e.g., $10 or $99 top-ups · Three plans offered: Free, Pay-as-you-go, and Enterprise · Pay-as-you-go tier carries a 5.5% platform fee, with fee discounts available · Small fee charged when purchasing credits; model inference pricing is passed through from
Platforms
Connects to 70+ providers (4 free providers and 25+ free models available)OpenRouter provides a models browsing/aggregation page
Adoption notes
DeploymentRuns at the edge with fallback to other providers when one goes down for higher availabili
LicenseNot disclosed by source
Model supportSupports latest model aliases (e.g., ~openai/gpt-latest) that always resolve to the newest · Krea 2 Large · Sparse mixture-of-experts language model from Meituan · Poolside Laguna S 2.1 · Gemini 3.6 Flash
Data controlCustom fine-grained data policies to restrict prompts to trusted models and providers · Workspace Guardrails enforce zero data retention, block prompt injection against 30+ OWASP
Learning curveIntermediate
Primary use casesMulti-model LLM orchestration with automatic provider fallbacks, Chat, image generation, embeddings, and transcription via one endpoint, Building agents with tool use, loops, and state management, Routing traffic across providers for higher availability, Accessing newest flagship models without code redeployment, Enterprise deployments with custom data policies and Private Models, Illustration, anime, painting, and other expressive artistic styles, Coding, repository-level changes, long-horizon problem solving, and agentic workflows, Software engineering and agentic coding use cases, Rapid iteration and graphic design exploration where fast generation is the priority
What to verify before adopting
- Free tier rate limit is 50 requests per day
Evolution and major updates
Research in progressOpenRouter timeline
A concise history of software releases and material product changes. Events appear only when a dated source supports what changed.
Scheduled for research
Building a reliable release history.
This profile is being checked for dated releases and material product changes. Nothing appears here until the exact date and event can be verified from a recorded source.
Dated event Short explanation Original source
Citation ledger
17 unique pagesRecorded sources
Facts, answers, structured details, milestones and primary resource links cite this shared ledger. Each external page appears once; release tags from the same GitHub project are grouped under one release history.
- 1openrouter.ai 13 facts · 4 answers · Official site
- 2openrouter.ai/pricing 6 facts · 3 answers · Pricing
- 3openrouter.ai/docs/quickstart 6 facts · 2 answers · Documentation
- 4openrouter.ai/support 6 facts · 2 answers · Official site
- 5openrouter.ai/blog 6 facts · 1 answer · Official site
- 6openrouter.ai/blog/announcements/may-release-spotlight 6 facts · 1 answer · Official site
- 7openrouter.ai/krea/krea-2-medium-turbo/pricing 6 facts · 1 answer · Pricing
- 8openrouter.ai/krea/krea-2-medium/pricing 6 facts · 1 answer · Pricing
- 9openrouter.ai/meituan/longcat-2.0/pricing 6 facts · 1 answer · Pricing
- 10openrouter.ai/poolside/laguna-s-2.1/pricing 6 facts · 1 answer · Pricing
- 11openrouter.ai/about 5 facts · 1 answer · Official site
- 12openrouter.ai/google/gemini-3.6-flash/pricing 6 facts · Pricing
- 13openrouter.ai/krea/krea-2-large/pricing 6 facts · Pricing
- 14openrouter.ai/thinkingmachines/inkling/pricing 6 facts · Pricing
- 15openrouter.ai/customers/paxhistoria 5 facts · Official site
- 16openrouter.ai/models 4 facts · 1 answer · Official site
- 17openrouter.ai/privacy Security
Research status98 substantive facts · 17 source pages · quality score 95/100
Connected directory
Semantic matchesExplore related solutions
These tools share a workflow, capability or audience with OpenRouter. They may complement it rather than replace it, and are not presented as integrations or endorsements.
