Hopper

Use this when the user needs a Hopper API key for their project. Creates a key on the signed-in account and returns its name, its first characters and a one-time link the user opens to copy it (the key itself is never in this result). Ask the user to store it as HOPPER_API_KEY in server-side secrets, such as a gitignored .env; don't ask them to paste it into the chat. Keys are managed at https://withhopper.com/console/keys.

create_api_key

Use this when the user asks about their Hopper credits, spend, usage or API keys. Shows the signed-in account: credits remaining, total spend, request and token counts, and API keys (names and first characters only; secrets are never readable). Read-only: it can't delete, revoke or reveal keys.

get_account

Use this when connecting a voice agent's LLM to Hopper. Returns Hopper's tested integration code for LiveKit Agents, Pipecat, Vapi, or the OpenAI SDK (for anything else). Each guide switches only the LLM, with an HTTP/2 client and a prompt-cache warm-up; STT, TTS and tools stay as they are.

get_integration_guide

Use this when the user asks which models Hopper serves, what they cost, or which model to use. Lists the LLM, text-to-speech and speech-to-text models with context length, prices (what the API bills) and whether each is available now.

list_models

Use this when choosing a voice for speak, or when the user asks what voices Hopper has. Lists the featured voices ready for text to speech, with id, name, language, gender and a short description. Works without an account.

list_voices

Use this when a voice agent responds slowly or before switching its LLM to Hopper. Checks the system prompt, tools and LLM client code for what adds time to first token: per-call data at the top of the prompt (breaks prompt caching), tools the prompt names but doesn't send (silent empty turns), reasoning_effort, no streaming, no HTTP/2, a client per request, no warm-up. Deterministic rules; nothing is sent anywhere or stored. Returns findings ranked by impact, each with the fix.

review_voice_agent_config

Use this when the user wants text read aloud or needs an audio file of speech. Shows an audio player and returns 24 kHz mono WAV audio with a playable HTTPS link (valid 10 minutes, 3 plays; download it to keep the file). Up to 4000 characters; longer text is split at sentence ends. Works without an account, limited to about 1 minute of audio per call and 10 minutes a day; when the user is signed in, usage is billed to their Hopper account without those limits. Do not use for music or sound effects.

speak

Use this when the user wants speech in an audio file turned into text, or the agent needs to hear a recording. English speech to text with word timestamps. Takes WAV, MP3, M4A, Ogg/Opus, FLAC or WebM (or raw PCM16 mono with sample_rate): a file the user attached (audio_file), a public https URL (audio_url, up to 25 MB), or base64 (audio_base64, up to 3 MB); up to 5 minutes of audio per call. For a local file, call it with no audio to get a one-time upload URL for curl (up to 4 MB). Works without an account, limited to 30 minutes of audio a day; when the user is signed in, usage is billed to their Hopper account.

transcribe