Orshot

Mint (or update) a public, signed render URL for a studio template and return it ready to use. Use this when you need a STABLE, reusable, cacheable image link — OG/social images, email images, docs/dashboard embeds — rather than a one-off render (orshot_generate_image). The returned URL renders the template on the fly; pass modification values as query params to produce variants from the same URL. Control access with `active`, `maxRenders`, `validTill`, and `cacheForSeconds`. Calling again for the same template updates only the fields you pass (partial update) and keeps the signature.

orshot_create_dynamic_url

Permanently delete a brand audio file.

orshot_delete_brand_audio

Permanently delete a saved brand color.

orshot_delete_brand_color

Permanently delete a custom font. Templates using this font will fall back to the default font.

orshot_delete_brand_font

Permanently delete a brand asset image.

orshot_delete_brand_image

Permanently delete a brand video.

orshot_delete_brand_video

Permanently delete a studio template. This action cannot be undone.

orshot_delete_template

Delete a workflow. This stops any future runs. Ask the user to confirm before deleting.

orshot_delete_workflow

Delete the dynamic URL for a template. The public URL stops working. Does not affect the template itself.

orshot_delete_dynamic_url

Delete a template folder. The folder itself is gone for good, but NO templates are deleted — every template inside returns to 'All templates' unfiled. Confirm with the user before calling this, since the grouping cannot be restored (re-filing each template is manual).

orshot_delete_template_folder

Permanently delete a draft or scheduled post, in Orshot and on the platforms it was queued with. Use it to cancel something scheduled or clear a draft the user rejected. This cannot be undone, and it only works while the post is still held — an already-published post cannot be taken down from here.

orshot_discard_social_post

Create a copy of an existing studio template. Returns the new template's ID.

orshot_duplicate_template

Explain which Orshot plan tier includes a feature or limit (workflows, embed, video, credits where 1 image or PDF page or video-second = 1 credit, members). Use this only to explain why a capability is unavailable on the workspace's current plan — compare against orshot_get_account. Do not initiate purchases, upgrades, or checkout, and do not quote prices; for plan details point the user to the informational pricing page in the response.

orshot_get_plans

Generate an image, PDF, or video from a studio template. Supports rendering to multiple formats with optional quality, scale, page selection, and video options. For multi-page templates, use includePages to select specific pages (e.g. [1, 3]). For PDF output, use pdfDpi to control print quality (default 72, use 300 for print). For video/GIF output, use videoOptions to control duration, fps, and trimming — plus soundtrack (audioSource swaps/adds page audio tracks; use brand audio URLs), burned-in subtitles (subtitleSource auto-transcribes any voiceover URL via Whisper, or takes .srt/.vtt; style with the subtitle* keys, karaoke via subtitleMode "word-by-word"), global muting, and multi-page stitching (combinePages + pageTransition). CHECK LAYOUT WITH A STILL, NOT A VIDEO. Before spending an mp4 (60-90s + video credits), verify composition with format:"png", includePages:[N], responseType:"base64" — about 1s, one image credit, and you SEE the frame. Layout, contrast, scrims, safe-area and text fit are all identical to the video's first frame; only motion is missing. Fix what you see, then render the video ONCE. Rendering an mp4 to find out whether a layer is in the right place is the most expensive mistake available here. LONG RENDERS: video/subtitle renders can take 60-90s. This server streams progress to keep the request alive; if your client times out anyway, the render still completes server-side (credits are used) — recover the output URL via orshot_list_workspace_logs instead of re-rendering. For guaranteed-no-timeout operation (or several long renders in a row), pass background: true — the tool returns a jobId immediately; collect it with orshot_get_render_job using waitSeconds: 25 (the server holds and polls for you; call again while finished is false). SMART RESIZE (one design → any size): the design is deterministically re-laid-out to fit a new canvas. - Use `size` to render the design at a DIFFERENT size (e.g. size: "instagram-story" or "1080x1080" or { width, height }). - Use `extraSizes` to ADD extra sized copies in the SAME call (e.g. extraSizes: ["instagram-story","1080x1080","1200x630"]) — the response nests an "extraSizes" array of { name, size, width, height, url } beside each page. Perfect when a user asks for one design across multiple social formats. RESPONSE SHAPE (so you surface the right links to the user): - Single-page: `data.data` is one object → `data.data.content` is the main output URL; `data.data.extraSizes` (if requested) is an array of variants, each `{ name, size, width, height, content }` where `content` is that variant's URL. - Multi-page: `data.data` is an array (one per page); each page has its own `content` and `extraSizes`. This tool also prepends a ready-to-share summary of every output + variant URL, so present those links to the user; a per-variant `error` means that one size failed (the others still succeeded). Optionally publish the rendered output to connected social accounts by including the "publish" parameter. IMPORTANT: Only include the "publish" parameter when the user explicitly asks to publish or post to social media. Never auto-publish. Use orshot_list_social_accounts first to get the account IDs needed for publish.accounts. 🃏 PREVIEW CARDS — the user sees ONE final card, not your build steps. Every call of this tool drops a preview card into the chat; a stack of superseded cards reads as clutter and confuses the user. Create the complete design with includeThumbnails:true so that first card can be final. Stop when its preview and checks are clean. Only when they reveal a concrete issue, batch the correction into ONE focused update; the update card then replaces the create card as the result. Multiple cards are fine ONLY when the user asked to see variants or before/after.

orshot_generate_image

Generate an image from a pre-built Orshot library template. Use orshot_list_library_templates to see available templates and orshot_get_library_template_modifications to see what parameters each accepts. Common templates: - website-screenshot: Capture a webpage (modifications: { websiteUrl, fullCapture?, delay?, width?, height? }) - tweet-image: Screenshot a tweet (modifications: { tweetUrl }) - beautify-screenshot: Add padding/shadow to a screenshot (modifications: { screenshotImage }) - open-graph-image-1: OG image for blog posts (modifications: { title, description }) - image-background-removal-api: Remove image background (modifications: { inputImage }) - dynamic-image-resizer: Resize/crop an image (modifications: { imageUrl, width?, height?, resizeType? }) 🃏 PREVIEW CARDS — the user sees ONE final card, not your build steps. Every call of this tool drops a preview card into the chat; a stack of superseded cards reads as clutter and confuses the user. Create the complete design with includeThumbnails:true so that first card can be final. Stop when its preview and checks are clean. Only when they reveal a concrete issue, batch the correction into ONE focused update; the update card then replaces the create card as the result. Multiple cards are fine ONLY when the user asked to see variants or before/after.

orshot_generate_image_from_library

Fetch the authenticated user's account: profile (name, email), workspaces, current plan, and live usage — credits, AI credits, voice seconds, member/workspace/rate limits, and which features (workflows, embed, video, audio) the plan includes. Call this to check whether a feature is available on the current plan. For the voice catalog and complete audio authoring limits, use orshot_get_audio_authoring_options. To explain which tier includes a feature, use orshot_get_plans.

orshot_get_account

Fetch the active workspace's details: title, slug, dashboard URL, creation date, and any saved onboarding answers (goals, cadence, destinations recorded by orshot_update_workspace). Use the onboarding block to pick up where a previous setup session left off instead of re-asking the user.

orshot_get_workspace

Everything measured for a single post: impressions, reach, likes, comments, shares, saves, clicks, engagement rate, per-platform results, and a day-by-day timeline since it published. Use it after orshot_get_social_analytics identifies a post worth understanding. A post published in the last few minutes will show no numbers yet — that is normal, not an error, so do not retry in a loop. Read-only, shares the 10 requests per minute analytics budget.

orshot_get_social_post_analytics

Call once before creating or updating a template with generated narration, ambient audio, sound effects, or automatic captions. Returns the account's current billing-period audio allowance and remaining seconds together with every supported speech model and the public voice catalog, so an agent can choose valid settings and author the complete video template without follow-up discovery calls. Usage is account-wide across workspaces; the active workspace is retained only for attribution. This tool does NOT generate audio. Put narration in pages_data[].audioTracks[].tts and ambient/SFX in pages_data[].audioTracks[].sfx when calling orshot_create_template_design or orshot_update_template_design. Saving a new recipe synthesizes a hosted track and consumes the account's billing-period voice allowance; an identical cached recipe does not consume it again. For dynamic personalized narration, set tts.parameterId and pass new spoken text through render modifications. Author narration at roughly 2 words per second with one second of headroom; create, update, patch, and dynamic video render tools preflight likely clipping before synthesis.

orshot_get_audio_authoring_options

When this workspace's audience actually engages, computed from its own posting history: the best day-and-hour slots (UTC), how posting frequency relates to engagement per platform, and how fast engagement decays after publishing. Use it before scheduling to pick a time, and prefer it over generic best-time advice because it is derived from these accounts. Slots need posting history to appear, so a new workspace can legitimately come back empty. Read-only, shares the 10 requests per minute analytics budget.

orshot_get_best_time_to_post

List audio files in the workspace's brand library. Use tag filtering to find tracks for specific purposes (e.g. "music", "voiceover", "sfx"). Tags use AND logic. Returns up to 100 audio files with metadata (duration, format, file size). Use a track's direct_url as the audio source on video template pages.

orshot_get_brand_audio

List saved brand colors in the workspace. Supports hex, rgb, hsl, and gradient types. Use tag filtering to find colors for specific purposes (e.g. "primary", "background", "accent"). Tags use AND logic. Returns up to 200 colors.

orshot_get_brand_colors

List custom fonts uploaded to the workspace. Use tag filtering to find fonts for specific purposes (e.g. "heading", "body", "brand"). Tags use AND logic. Returns up to 100 fonts. When creating templates, use the font's `name` field (without extension) as the fontFamily in element styles. Custom fonts take priority over Google Fonts.

orshot_get_brand_fonts

List brand asset images in the workspace. Use tag filtering to find assets for specific purposes (e.g. "logo", "background", "product"). Tags use AND logic — all specified tags must match. Returns up to 100 images, newest first.

orshot_get_brand_images

Fetch all brand assets at once — images, colors, fonts, videos, and audio. Optionally filter each type by tags. This is the recommended starting point when creating branded templates. Use tag filtering to get only relevant assets for your design task. Example workflow: 1. orshot_get_brand_kit with tags matching your design purpose (e.g. "social-media") 2. Use returned brand colors for backgrounds, text colors 3. Use returned brand fonts for text elements (set fontFamily to the font name without extension) 4. Use returned brand images for logos, backgrounds, product shots 5. Use returned brand videos for video elements 6. Use returned brand audio for background music / voiceover on video templates

orshot_get_brand_kit

List brand videos in the workspace. Use tag filtering to find videos for specific purposes (e.g. "intro", "background", "product"). Tags use AND logic. Returns up to 100 videos with metadata (width, height, duration, format).

orshot_get_brand_videos

Day-by-day social metrics for this workspace, plus a per-platform breakdown of the whole window. Use it for trends over time ("are impressions going up?", "how did last week compare"), not for individual posts — orshot_get_social_analytics covers those. Read-only, and shares the 10 requests per minute analytics budget.

orshot_get_social_analytics_daily

Fetch the active workspace's Orshot Embed (white-label editor) configuration: whether it's set up, allowed domains, branding (title, icon, accent color), feature toggles, webhook, custom buttons, view count, and — importantly — whether the CURRENT PLAN allows embed and which embed features it includes. The signing secret is never returned through the API; the response includes the dashboard URL where the user can view it.

orshot_get_embed_config

Get the available modification parameters for a specific library template. Use this to discover what fields you can customize (e.g. websiteUrl for website-screenshot, tweetUrl for tweet-image).

orshot_get_library_template_modifications

Check the status of an async render job (video, PDF, or image) started in the background. Returns status (queued | processing | succeeded | failed | canceled), finished (true once terminal), and on success the output URLs. Pass waitSeconds: 25 and the SERVER holds the call, checking the job until it finishes or the wait elapses — then simply call again while finished is false (typically 1-6 calls for a video). Do not try to wait between calls yourself; agents have no sleep primitive, which is exactly what waitSeconds replaces. Job records expire 7 days after creation; the rendered file itself is unaffected by job expiry. Do NOT re-render while a job is queued or processing — that double-spends credits.

orshot_get_render_job

How this workspace's social posts performed: impressions, engagement, clicks and engagement rate per post, plus follower counts per account. Start here for "how are we doing", "what worked" or "which post did best" — sort by engagement to rank winners. Covers posts synced from the platforms too, not just ones published through Orshot, so there is history from day one. Results are paginated: the reply states the workspace total and whether more pages exist, so quote the stated total rather than counting the rows you can see. Read-only. Analytics allows 10 requests per minute per workspace and responses are cached, so ask once and reuse the answer rather than polling.

orshot_get_social_analytics

Follower counts and growth for each connected social account, with a day-by-day follower history. Use it for "how is our audience growing", "which account grew this month" or to compare accounts. For post performance use orshot_get_social_analytics instead. Read-only, shares the 10 requests per minute analytics budget.

orshot_get_social_followers

Get full details of a studio template including design structure. Shows the template's stored page thumbnails as an inline preview card (multipage templates get a swipeable carousel) — use this when the user asks what a template looks like. 🃏 PREVIEW CARDS — the user sees ONE final card, not your build steps. Every call of this tool drops a preview card into the chat; a stack of superseded cards reads as clutter and confuses the user. Create the complete design with includeThumbnails:true so that first card can be final. Stop when its preview and checks are clean. Only when they reveal a concrete issue, batch the correction into ONE focused update; the update card then replaces the create card as the result. Multiple cards are fine ONLY when the user asked to see variants or before/after.

orshot_get_studio_template

Get the available modification parameters for a studio template. Returns the list of dynamic fields (text, image, etc.) that can be customized when rendering via orshot_generate_image. The response contains a "modifications" array where each entry has: - id: The key to use in the modifications object when rendering. For single-page templates this is just the parameterId (e.g. "headline"). For multi-page templates it is prefixed with the page number (e.g. "page1@headline", "page2@cta_text"). - type: The parameter type — "text", "imageUrl", "videoUrl", "backgroundColor", "fill", "color" - description: Human-readable label Beyond these listed keys, EVERY parameterizable element also accepts implicit dot-notation render props (not listed in the response): - Style: "paramId.fontSize": "64px", "paramId.color": "#FF0000", etc. - Animation timing (video renders): "paramId.showAt", "paramId.hideAt", "paramId.enterType"/"enterDuration"/"enterEasing"/"enterReverse", "paramId.exitType"/"exitDuration"/"exitEasing"/"exitReverse", "paramId.ambientType"/"ambientDuration"/"ambientEasing" - Video elements: "paramId.trimStart", "paramId.trimEnd", "paramId.muted", "paramId.loop" - Generated narration: a track with tts.parameterId is listed here with type "voiceText"; pass new spoken copy under that id. Existing soundtrack-file swaps use videoOptions.audioSource. There is no generic audio parameterType. - element: The element details (type, current default content, parameterId) Also returns "modifications_json" — a flat key→defaultValue map you can use as a starting point and override specific keys.

orshot_get_studio_template_modifications

Get a studio template's public sharing state — whether it's shared, the public link, the embed page + ready-to-paste iframe code, whether visitors can duplicate it, and view/copy stats. Use orshot_update_template_sharing to change any of it.

orshot_get_template_sharing

Get a workflow's full definition — status, trigger, and its ordered steps (with config). In clients that support inline previews this also renders the workflow as a visual card (node pipeline with icons), so ALWAYS call this when the user asks to see/show a workflow — never build your own visualization of it.

orshot_get_workflow

Get one run's detail including per-node results — which step ran, its status, output, and any error. Use this to debug why a workflow run failed.

orshot_get_workflow_run

Get a workflow's public sharing state — whether it's shared, the public link and embed page, and whether its template is included when someone copies it. Use orshot_update_workflow_sharing to change any of it.

orshot_get_workflow_sharing

Get (or create) the inbound webhook URL for a webhook-triggered workflow — the URL to POST events to from your own app or a tool like Zapier/Make, plus its signing secret. Only relevant for workflows whose trigger is a webhook.

orshot_get_workflow_webhook

Read one post in full: caption, media, and per-account status with live post URLs or failure reasons. This is how you confirm what actually happened to a post — including confirming that a draft is still held and has NOT been published. Posts still in flight are refreshed from the platforms first, so there is no separate sync step. In clients that support inline previews this also renders the post as a visual card (media + caption + status), so call this when the user asks to see/show a post.

orshot_get_social_post