Test any LLM API key in seconds
Paste a key, list your accessible models, benchmark latency, and copy a snippet. Works with 17 providers. No signup.
Stateless proxy — keys never logged, stored, or persisted. What happens to your key →
Keys never persist. The proxy forwards once and forgets.
p50 / p95 / p99 from a 5-ping benchmark, per provider.
cURL, Python, Node fetch, axios — copy and ship.
Providers
Pick a provider, paste a key, ship.
Anthropic
/test/anthropic-api-key
Claude Opus, Sonnet, Haiku.
chatvisionOpenRouter
/test/openrouter-api-key
One API, every model.
chatvisionopen-sourceGroq
/test/groq-api-key
LPU-accelerated Llama, Mixtral, Whisper.
chataudioPerplexity
/test/perplexity-api-key
Sonar online search-augmented chat.
chatxAI
/test/xai-api-key
Grok-2 and Grok-2-vision.
chatvisionFireworks AI
/test/fireworks-api-key
Fast inference for OSS models.
chatembeddingopen-sourceHugging Face
/test/huggingface-api-key
Inference API + Inference Endpoints.
chatembeddingimageaudioopen-sourceReplicate
/test/replicate-api-key
Run open-source models with one HTTP call.
chatimageaudioopen-sourceOpenAI
/test/openai-api-key
GPT-4o, o1, embeddings, DALL·E.
chatembeddingimageaudiovisionDeepSeek
/test/deepseek-api-key
DeepSeek-V3 chat and coder.
chatGoogle Gemini
/test/gemini-api-key
Gemini 2.0 Pro, Flash, embeddings.
chatembeddingvisionCohere
/test/cohere-api-key
Command-R chat, rerank, embed.
chatembeddingMistral
/test/mistral-api-key
Mistral Large, Codestral, embeddings.
chatembeddingTogether AI
/test/together-api-key
200+ open-source models, fast.
chatembeddingimageopen-sourceAzure OpenAI
/test/azure-openai-api-key
OpenAI models on Azure infrastructure.
chatembeddingvisionAWS Bedrock
/test/bedrock-api-key
Anthropic, Meta, Cohere, Mistral on AWS.
chatembeddingimageOllama
/test/ollama-api-key
Run open-source models locally, no key.
chatembeddingopen-source
FAQ
What KeyForge is — and isn't.
- What is KeyForge?
- KeyForge is a free, open-source developer tool that tests Large Language Model (LLM) API keys. Paste a key from OpenAI, Anthropic, Google Gemini, Groq, Mistral, Cohere, DeepSeek, xAI, Perplexity, Together AI, Fireworks AI, OpenRouter, Azure OpenAI, AWS Bedrock, Hugging Face, Replicate, or a local Ollama host, and KeyForge verifies the key works, lists every model it can call, benchmarks request latency, and generates ready-to-use cURL, Python, and Node.js snippets.
- Is KeyForge related to the KeyForge card game?
- No. This site is a developer tool for testing LLM API keys. It is unrelated to the trading card game KeyForge published by Asmodee / Fantasy Flight Games. The naming overlap is coincidental.
- What does KeyForge do with my API key?
- Nothing — beyond forwarding it once to the upstream provider so we can read the response. The proxy is stateless: keys are never logged, stored, or persisted. There is no database, no analytics on key values, and no third-party scripts. The full code is open source on GitHub.
- Which LLM providers does KeyForge support?
- OpenAI, Anthropic (Claude), Google Gemini, Mistral, Cohere, Groq, DeepSeek, xAI (Grok), Perplexity, Together AI, Fireworks AI, OpenRouter, Azure OpenAI, AWS Bedrock, Hugging Face, Replicate, and Ollama (local). Adding a new provider takes about 30 minutes — see the contributing guide.
- Do I need to sign up?
- No. KeyForge is free and requires no account, email, or signup. Just paste a key and test.
- Can I run KeyForge myself?
- Yes. The codebase is MIT-licensed and runs on any Node.js host or Vercel. Clone the repo at github.com/dimssu/key-forge and follow the README.
Guides
Practical, opinionated, no fluff.
/guides/how-to-get-an-openai-api-key
How to get an OpenAI API key
Step-by-step walkthrough to create, scope, and verify an OpenAI API key — including project keys and service accounts.
Read/guides/how-to-get-an-anthropic-api-key
How to get an Anthropic API key
Create a Claude API key from the Anthropic console, set workspace limits, and verify it in seconds.
Read/guides/how-to-get-a-gemini-api-key
How to get a Google Gemini API key
Generate a Gemini API key from Google AI Studio, switch to a billed project, and avoid the free-tier surprises.
Read/guides/api-key-security-best-practices
API key security best practices for LLMs
How to store, scope, rotate, and revoke LLM API keys without leaking them through git, logs, or shared environments.
Read