Skip to content
9 providers · 1 is free with no key

One prompt. Every open inference provider.

Send the same prompt to Groq, Together, OpenRouter, Hugging Face and more, side by side. See the output, time to first token and tokens per second before you commit to a provider.

Try the playground Browse providers
Glowing server nodes connected by light paths that converge on a single point
Example runtok/s
  • Provider A0.21s · 412
  • Provider B0.18s · 590
  • Provider C0.44s · 118

Illustrative numbers. Run your own in the playground.

Works with OpenAI-compatible endpoints from

Built for an honest comparison

Same prompt, same parameters, same moment. The differences you see come from the provider, not the setup.

Three glass panels with particle streams moving at different speeds

Up to three providers in parallel

Requests fire at the same time and stream into their own panel, so latency differences are visible as they happen.

A translucent cube unfolding into layered grids of glowing points

Open models first

Llama, Qwen, DeepSeek, Mistral and GPT-OSS. Pick a suggested model or type any model ID the provider serves.

From zero to a comparison in under a minute

  1. 01

    Pick providers

    Choose up to three. Pollinations works instantly; others take your API key.

  2. 02

    Write a prompt

    Add an optional system prompt, set temperature and max tokens once for all.

  3. 03

    Compare

    Read the answers side by side with latency and speed under each one.

curl https://api.groq.com/openai/v1/chat/completions \
  -H "Authorization: Bearer $GROQ_API_KEY" \
  -d '{
    "model": "llama-3.3-70b-versatile",
    "messages": [{"role": "user", "content": "Hi"}]
  }'

Stop guessing which provider is fastest.

No account needed. Start with the free provider, add your own keys when you want more.

Open the playground