# Ask anything, launch everything — introducing Ask in the eroq studio > A product note on Ask, the bar on the eroq studio home. Type or say what you want and it sets up the right tool, finds your work or answers — no credits spent. Published 2026-09-29 · eroq.ai — canonical: https://eroq.ai/blog/ask-anything-launch-everything The studio has a lot of doors — Video, Image, Voice, Text, Cinema, Replace, the library, your characters, billing, API keys — and each tool has its own page and its own settings to learn. **Ask** takes a sentence instead. Say what you want and the studio sets itself up for it: the right tool, the settings you named, a prompt written from your words. The Generate click stays yours. This is the product note: where Ask lives, the six things it does, and what it deliberately leaves alone. The short version is on the [Ask page](/ask). ## Where it lives Ask is the bar at the top of [the studio](/studio) home, after sign-in. It is the composer's default tab; its placeholder says what it is for: **"Ask anything, launch everything"**. - **Type, then press Enter.** Shift+Enter adds a new line. - **Or tap the mic and speak.** Ask uses the browser's own speech recognition, in Chrome, Edge and Safari; in browsers without it, such as Firefox, the mic is hidden. When you stop talking, your words send themselves. - **Or press ⌘K** (Ctrl+K on Windows) on any studio page. The topbar search reads **"Search or ask anything…"** — type a phrase, and the last row, **"Ask eroq: “…”"**, sends it to your chat in progress, or starts one. The first message turns the home into a conversation: your message on the right, the reply on the left, written out word by word. While Ask works, a small orb turns and short hints show where it is — "Reading that…", "Picking the right tool…", "Setting the parameters…". ## The six things it does In each example, the first line is what you type; the line after the arrow is what appears under Ask's reply. ### 1. It sets up a tool Ask opens [Video](/studio/video), [Image](/studio/image), [Voice](/studio/voice), [Text](/studio/text), [Cinema](/studio/cinema) (films) or Replace (video-to-video edits of your own clip) and fills in what you asked for: the engine you name, the clip length, the aspect ratio, the resolution, the camera move (one of 24 — push-in, tracking, orbit, dolly zoom, FPV…), a palette or lighting look, the number of takes (1–4), the characters and elements you name, the voice (roster voices Aria and Orion, your cloned voices, or a character's voice) and the library folder to save into. Then it writes the prompt. ```text make a 10-second vertical clip with Kling of a woman walking through neon rain, slow push-in → Setting up Video · Kling 3.0 · 10 s · 9:16 with your parameters… ``` The rule that matters: **it only sets what you asked for.** Anything you did not say stays as the tool already has it. The [step-by-step guide](/blog/make-an-ai-video-by-describing-it) takes one sentence to a rendered clip. ### 2. It says what it does, in plain sight That grey line is deliberate. It names the tool and the settings it filled — engine, length, format — before anything happens, and about a second later the tool page opens with everything filled in. When Ask only *proposes* — you asked a question, the request was ambiguous, or you named an engine your plan does not include — you get a button instead of an automatic jump, and a plan badge next to a locked engine. On a plan without Veo 3.1: ```text make a clip of a perfume bottle on black marble, slow orbit, on Veo 3.1 → Video · Veo 3.1 [Open Video →] Creator badge ``` Click through and the page opens with the engine's Unlock button. ### 3. It goes anywhere in the studio "Open my library", "billing", "clone my voice", "my API keys", "show me what the community made", "new film" — no menu to remember. ```text show me what the community made → Opening Explore… ``` ### 4. It finds your work by description You remember the dragon, not the date. Ask searches your creations by the words of their prompt or title and shows up to five matches as thumbnails; if nothing matches, you get the [library](/studio/library) filtered on those words. Other languages work too — "ouvre la vidéo du dragon" runs the same search. ```text open the dragon clip → 3 matches for “dragon” ``` Characters, elements and cloned voices are found by name. More in [finding old renders by description](/blog/find-old-ai-renders-by-description). ### 5. It answers questions "How many credits do I have?", "what's the cheapest video engine?", "which engine does 1080p with sound?" — Ask answers from the studio's live catalogue and offers to open the right tool. ```text which engine does 1080p with sound? → an answer from the live catalogue, then [Open Video →] to try it ``` ### 6. It asks back When the medium itself is unclear, Ask asks one short question instead of guessing, with the answers as chips: ```text make something cool → An image, a clip, a voice line or some text? [Image] [Video] [Voice] [Text] ``` ## It never spends credits Ask prepares and opens; it does not render. The **Generate** button on the tool page stays your click, with the usual credit check and plan unlock. A 5-second clip on [Seedance 2.0 Mini](/models/seedance-2-0-mini) — the engine a new free account starts on — costs 45 credits at that click, and nothing before it. **Asking itself is free.** Ask, change your mind, ask again — nothing is charged before Generate. ## Any language, typed or spoken Write or speak in your own language, and Ask answers in it. Two rules decide what lands on the tool page: - **Video and image prompts are written in English**, for the engines. Ask in Spanish and the prompt on the Video page is English. - **A voice line keeps the language you wrote it in.** It is spoken as written. [Voice commands for AI video](/blog/voice-commands-for-ai-video) covers the mic; [AI video prompts in any language](/blog/ai-video-prompts-in-any-language) covers the language side. ## Follow-ups keep the thread The conversation keeps context, so the next message can be short: ```text the same but vertical rather with Kling make it 5 seconds now as an image ``` What you said earlier still counts: "rather with Kling" keeps the subject, the length and the format you already gave. The [conversation guide](/blog/iterate-ai-video-by-conversation) has more. Each conversation is a chat of its own, listed newest first in a column beside the chat (on a phone, behind **Chats**). After Ask opens a tool, your browser's back button returns to the chat, the **Ask** row in the studio's left rail takes you back to the one in progress, and any chat in the list picks up where you left it. **New chat** starts a fresh one and keeps the last in the list. Chats are kept in this browser for 30 days and are not synced across devices. ## Films and cast "A short film…" opens [Cinema](/studio/cinema) with a new film whose scene one is your description; ask for several scenes and describe them, and each one gets its own scene card, up to eight. Name one of the six recipes — Perfume ad, Music video opener, Horror teaser, Streetwear drop, Romance short, Travel diary — and the whole recipe opens as a script. Writing a film costs nothing; each scene is billed when you render it. [From one sentence to an AI short film](/blog/from-one-sentence-to-an-ai-short-film) walks through it. [Characters](/studio/characters) join by name, with one condition: a character appears in renders only once it has its **look sheet**, made in its Identity tab; one without is flagged "Needs a look sheet". Places and objects work the same way; styles need no sheet. ## What Ask deliberately does not do - **Render by itself.** Every render starts with your click on the tool page. - **See inside your media.** It finds your work by prompt and title, not by looking at the pictures. - **Pretend.** eroq has no upscaling, no lip-sync and no inpainting, and Ask says so instead of opening a tool that can't do it. - **Act as an API endpoint.** Ask is part of the studio. Developers call the generation endpoints directly, or connect assistants through the [MCP server](/mcp). ## FAQ ### Does Ask cost credits? No. Asking is free, and Ask never spends credits: it prepares a tool page and opens it. Nothing renders and nothing is charged until you click Generate on that page, with the same credit check and plan unlock as always. ### Why did Ask show a button instead of opening the tool? It was proposing, not launching: you asked a question, the request was ambiguous, or you named an engine your plan does not include. In that last case a plan badge sits next to the button, and the page opens with the engine's Unlock button. ### Which languages does Ask understand? Write or speak in your own language and Ask answers in that language. Video and image prompts are written in English for the engines; a voice line keeps the language you wrote it in. ### Why can't I see the mic? The mic uses the browser's own speech recognition, which Chrome, Edge and Safari have. In browsers without it, such as Firefox, it is hidden; typing works everywhere. ### Can I use Ask through the API? No. Ask is part of the studio, not an endpoint. Developers call the generation endpoints directly, or connect an assistant through the MCP server at [/mcp](/mcp). [Open the studio](/studio) and type the first thing you want to make — asking is free.