Hope
chat & voice assistant

AI MixMasters

Per-source templates: 808, vocals, drums, master, FX returns. Tap the floating mic to start a live call with Hope — she reads and edits the chain in real time.
How to use: click "+ Add plugin" on any bus, search, click to add. ↑/↓ to reorder, × to remove. Bring the chain into a chat with Claude in Hope's chat, then hit Snapshot → Claude chat to save the discussion as a Session Snapshot pill next to START CALL.

Conversation

Hope (voice) and typed Claude messages share one transcript. Type for text-only Claude with full workbench context; press Start call for a live voice session with Hope.
Start a conversation. Type a message below for text-only Claude with full workbench context, or press Start call for a live voice session with Hope. Both flows render here.
 

Snapshot pills

Favourited chats from Hope's chat — click to recall. Mid-call: sent to Hope as context. Idle: drops into your compose box as a draft.
Hover a pill: apply · unfavourite · edit · copy · × delete

Session Snapshots

Every snapshot you copy to chat is auto-saved here. Latest first. Tap an entry to expand it.
No snapshots yet — chat with Claude in Hope's chat and hit Snapshot → Claude chat to save the discussion. Saved entries auto-favourite as pills above.

Knowledge Base

Paste research from NotebookLM, YouTube transcripts, articles, or interview notes. Active notes are injected into Claude's prompt every chat & voice session.
0 / 8000 bytes No notes yet.

Hope's memory

Cross-conversation profile of you. Auto-extracted at end-of-call via Haiku, hand-editable. Injected into every Hope conversation so she remembers across sessions.
0 / 2000 bytes
Mix Check
Audio Specs
Loudness
– –
Peak
– –
DR
– –
RMS– –
Crest factor– –
LRA– –
Phase / correlation– –
Stereo width– –
Sample rate– –
LUFS short-term– –
True peak– –
Headroom– –
Noise floor– –
Dissonancewith full analysis
Tempo
– –
Key
– –
approx
Classified
Genre– –
Subgenrewith full analysis
Production stylewith full analysis
Energywith full analysis
Moodwith full analysis
Input
Drop a WAV here
▾ browse file · live input · capture tab
0:000:00
or meter live: idle — readings update while it listens; stop to lock them in
Spectral Balance
drop a file to analyse
LowLow-MidHigh-MidHigh
Low 20–250 Hz
– – dB
Mid 250 Hz–3 kHz
– – dB
High 4–20 kHz
– – dB

Spectral Balance (real Tonal-Balance FFT — legacy view)

0
-6
-12
-18
-24
-36
SUB
BASS
LOW MID
MID
HIGH MID
AIR
20501002005001k2k5k10k20k
Stereo Balance
– –
Overall Balance
– –
Correlation
– –

Troubleshooter

Click any symptom for a recipe. Pick multiple — they stack on the snapshot.
Symptoms below feed the same diagnosis Claude sees on the next snapshot.

Platform Loudness Comparison

Platform
Target
Your Mix
Status
Competitive TrapClub / DJ play
-8 LUFS
– –
awaiting file
Spotify / YouTubeNormalised to -14
-14 LUFS
– –
awaiting file
Apple MusicSound Check on
-16 LUFS
– –
awaiting file
BroadcastEBU R128
-23 LUFS
– –
awaiting file

True Peak Ceilings

Destination
Ceiling
Your Mix
Status
StreamingMP3 / AAC safe
-1.0 dBTP
– –
awaiting file
SoundCloudStricter encoder
-0.5 dBTP
– –
awaiting file
Safe HeadroomStems / print masters
-2.0 dBTP
– –
awaiting file
Hard CeilingNever exceed
0 dBTP
– –
awaiting file
Current genre: Trap — plugins ⭐ = top pick for this genre  •  ❤ = your favorite

⚙ Settings

Keys, models, costs, session safety. Hope's chat (the rail, docked on every tab) handles the actual conversation; this is the plumbing.

API keys

No key saved. Optional — without it the AI falls back to its built-in (sometimes shaky) knowledge.
No key saved. Required — Hope's chat needs your EL key + Agent ID.
No agent ID saved.
Not yet registered this session. Safe to re-run — reuses the existing tool if it's already there.
No agent ID saved.
No agent ID saved.
No agent ID saved.
No agent ID saved.
No agent ID saved.

Models

Costs

ElevenLabs Conversational AI
Balance— set —
Total credits— update —
Used— update —
Remaining
Anthropic Research
Balance— set —
Session$0.000
Spent$0.000
Left
Typed chat — this session
Session: $0.000 Last reply: Messages: 0
Typed messages cost Anthropic credits. Voice (Hope) is billed via your ElevenLabs plan.

Session safety

Session tools cost & safety

Call duration
EL rate: $0.08/min · 275 min included/mo
Session time nudge
Amber pulse when call exceeds limit.
Session breakdown
Voice (duration)
Research (Anthropic)$0.000
Session breakdown — no previous session
Voice (duration)
Research (Anthropic)
Auto-pause if stuck
Saves you from a runaway session if the model hangs.
Usage dashboards
Real numbers live on the providers' sites.

Notes

How it works: your ElevenLabs key + Agent ID open a live conversation with Hope — ElevenLabs voice, Claude Sonnet 4.6 brain. She has tools for your chain, library, meters, insight notes, and flags, so she reads and edits the workbench while you talk. Try: "what's on my master bus?", "add Maag EQ4 after Ozone", "switch genre to R&B", or "I'm hearing pumping on the master."  At end of call your profile auto-extracts via Haiku so she remembers across sessions. Live transcript and snapshots show in Hope's chat, docked in the rail on every tab.

Reset

This wipes your entire workbench — chain, meters, flags, favourites, and custom plugins. It cannot be undone.

Marketing

Strategy, web presence, release planning, and future direction — powered by your Pop A&R persona.
Coming soon. This tab will connect to your Pop A&R voice persona — discuss AIMixmasters.com, marketing strategy, artist positioning, release planning, and where you want to take the project next. The persona system (Backlog #1) wires this up. In the meantime, bring these topics into the Hope tab.

Community

Social presence, community inbox, GitHub, and fan engagement — all in one place.
Coming soon. This tab will be the hub for the AI Mix Masters community — monitor social reactions, manage a community inbox for ideas and feedback, link to the GitHub repo and community pages, and engage with fans and collaborators. Scope is on the roadmap; the tab is here as a placeholder so it stays top of mind.
Copied