Chat completion

Low-latency chat completion on Groq LPUs.

needs keyunknown~2s typical

llmgroqllamafast

$0.0010

per 1k tokens + $0.001 call

plus $0.001 platform fee per run

Bring your own Groq key: provider portion drops to $0.

Not enabled on this deployment yet. Connect your own key to use it now.

Provider

$Set up https://cracked-ai-pearl.vercel.app/SKILL.md, then use Cracked to run groq /chat for me.

Input

Field	Type	Description
prompt*	string	User prompt
system	string	Optional system prompt
model	string	Model id default "llama-3.3-70b-versatile"
max_tokens	number	Max output tokens default 1024
temperature	number	Sampling temperature default 0.7
json	boolean	Ask for a JSON object response default false

Call it

curl

curl https://cracked-ai-pearl.vercel.app/v1/run \
  -H "Authorization: Bearer ck_live_..." -H "content-type: application/json" \
  -d '{"provider":"groq","endpoint":"/chat","input":{"prompt":"...","model":"llama-3.3-70b-versatile","max_tokens":1024,"temperature":0.7,"json":false}}'

cli

npx cracked-ai run -p groq -e /chat -i '{"prompt":"...","model":"llama-3.3-70b-versatile","max_tokens":1024,"temperature":0.7,"json":false}'

mcp

run_tool({ provider: "groq", endpoint: "/chat", input: {"prompt":"...","model":"llama-3.3-70b-versatile","max_tokens":1024,"temperature":0.7,"json":false} })

Try it

Runs against your signed-in workspace balance. Sign in if you have not.

Provider not enabled: the run returns a 503 with instructions.