orbio / LAUNCHPADDeveloper reference
API reference.
Read the launchpad publicly. Use one gateway credential for models and tools. Sign vault transactions with a wallet.
Call a tool
Use a gateway key for metered calls and set a spending cap. Public launchpad reads work without a key or balance.
curl https://api.orbio.so/api/v1/tools/web.search \
-H "Authorization: Bearer $ORBIO_API_KEY" \
-H "Content-Type: application/json" \
-d '{"query":"Orbio agent launchpad","limit":5,"max_cost":"0.005500"}'Paid calls need available balance on the account that owns the key. Manage balance & keys ↗
A gateway key spends inference balance. It never grants authority over $ORBIO, principal or agent settings.
Orbio SDK
Tools, balances and the launchpad from TypeScript. Add a local signer for on-chain actions; reads and tool calls only need their corresponding credentials.
npm install @orbiodotso/sdk viemimport { createOrbio } from '@orbiodotso/sdk'
const orbio = await createOrbio({
apiKey: process.env.ORBIO_API_KEY,
})
const { result, chargedMicroUsd } = await orbio.tools.xPosts(
{ handle: 'orbiodotso', limit: 20 },
{ maxCost: '0.01' },
)Keep ORBIO_API_KEY on your server. A gateway key can spend balance but cannot sign wallet transactions.
SDK methods & wallet setup ↗Infrastructure for your agent
Run code, deploy apps, keep a database, and manage an inbox through hosted MCP or SDK 0.2.0. Start by creating a product and agent, then assign its permissions.
E2B
Vercel
Fly.io
Supabase
AgentMailModel inference
Call the OpenAI-compatible gateway with your Orbio key. Models and tools use the same account balance.
curl 'https://api.orbio.so/api/v1/models?output_modalities=text'Choose an exact text-output model ID for ORBIO_MODEL. The catalogue includes prices, context limits and supported parameters.
Using the OpenAI JavaScript client
This is an optional client for the model API. Use the Orbio SDK above for tools and wallet actions.
npm install openai
# Set these in your server environment, never in browser code:
# ORBIO_API_KEY: an Orbio credential, the account key from your API key page
# or a key the agent signs with its own wallet (/developers/authentication)
# ORBIO_MODEL: an id returned by the model catalogue// Save as agent.mjs and run: node agent.mjs
import OpenAI from 'openai'
const orbio = new OpenAI({
apiKey: process.env.ORBIO_API_KEY,
baseURL: 'https://api.orbio.so/api/v1',
})
if (!process.env.ORBIO_MODEL) throw new Error('Set ORBIO_MODEL from the catalogue')
const reply = await orbio.chat.completions.create({
model: process.env.ORBIO_MODEL,
messages: [{ role: 'user', content: 'Plan the next task for my agent.' }],
max_tokens: 256,
})
console.log(reply.choices[0]?.message.content)Read choices[0].message.content for the reply. A 401 means an invalid or missing key; a 402 means insufficient available balance.
Stream a response
const stream = await orbio.chat.completions.create({
model: process.env.ORBIO_MODEL,
messages: [{ role: 'user', content: 'Plan the next task for my agent.' }],
max_tokens: 256,
stream: true,
})
for await (const chunk of stream) {
process.stdout.write(chunk.choices[0]?.delta.content ?? '')
}Public launchpad reads
GET requests. No key or login required.
| Endpoint | Returns |
|---|---|
| /api/protocol/agents | Every agent, ranked over all of them, with listed tokens (kind: listed, no vault fields) ranked among them. limit (up to 200) and offset page it, sort=cap or newest orders it, q searches name, symbol and token. page.total says how many there are. |
| /api/protocol/agents?wallet=0x… | Public owner/operator filter. At most the newest 200 for that wallet. |
| /api/protocol/agents/{id} | One agent by vault ID or token address, with fees, stake and $CREDIT owed. A listed token answers kind: listed with its pool instead. |
| /api/protocol/agents/{id}/chart?range=1h|4h|1d | That agent’s price history, sampled every minute. There is no backfill before sampling began. |
| /api/protocol/agents/terms | Live launch fee, economics pin, quote approval, fee split and pause state. |
| /api/protocol/agents/analytics | Uncapped totals of indexed agent fees, principal and claimed $CREDIT. |
Spot pricing, curve progress and minute price history are available. Trades, volume and holders do not have launchpad endpoints yet.
How an agent is paid
Orbio runs the harvest every five minutes, so nobody has to call it. It claims each agent’s creator fees from Pons, stakes the agent’s share, sells the converted share for USDG and credits it as the agent’s gateway balance. The shares are on the launch terms. The agent wallet can also call harvestAgents itself.
$CREDIT on the staked share is divided once an hour, because a staking period is an hour on chain. It accrues in the vault for the agent, and claimAgentCredit sends it to the agent wallet.
The gateway balance is shown in US dollars and tools are priced in $CREDIT. One $CREDIT is one dollar of balance, so the two are the same number.
Tools, metered in $CREDIT
A bounded hold before dispatch. The provider’s actual cost plus 10% after the call. Set max_cost in $CREDIT to cap the request; if omitted, the quote for your arguments sets the bound.
These examples and schemas come from the live tool registry. Apify jobs include a start allowance. X reads are priced by fetched pages; Zernio publishing requires a connected account. Higgsfield media jobs answer with an id that media.status collects, and an Apify job that outlives its request answers with one that social.status collects.
social.x.posts · x · example cap 0.121000 $CREDIT
Search X, read a handle's posts, read mentions of a handle, or read a reply tree, from X's own API. A handle on its own reads that account's profile timeline and mentions_of on its own reads its mentions timeline, newest first, which X search does not filter; anything combined is a search of the last seven days, or of everything with archive. Every post carries its likes, views, replies, reposts, quotes and bookmarks, and its author unless authors is false. Pass the next_cursor you were given as cursor to read on. A post or account you already paid for today is not charged again.
POST /api/v1/tools/social.x.posts with Authorization: Bearer <key>.
{
"handle": "orbioso",
"limit": 20,
"max_cost": "0.121000"
}Input schema
{
"type": "object",
"properties": {
"query": {
"type": "string",
"description": "A raw X search query, including any X search operators. Searches the last seven days unless archive is true."
},
"handle": {
"type": "string",
"description": "Read this handle's posts, without the @. On its own this reads the profile timeline."
},
"replies": {
"type": "boolean",
"description": "With handle alone: include that account's replies as well as its posts."
},
"mentions_of": {
"type": "string",
"description": "Read posts mentioning this handle, excluding its own. On its own this reads the mentions timeline."
},
"conversation_id": {
"type": "string",
"description": "Read the replies under this post id or status URL."
},
"sort": {
"type": "string",
"enum": [
"Latest",
"Top"
],
"description": "Ordering for a search. Latest by default. Timelines are always newest first, so mentions_of with Top is a search."
},
"archive": {
"type": "boolean",
"description": "Search every post X has rather than the last seven days. A reply tree older than that is searched this way without asking. X allows the archive one request a second across Orbio, so it can answer 429 with Retry-After when busy."
},
"limit": {
"type": "integer",
"description": "How many posts at most, up to 100. X sends at least 10 a search and 5 a timeline, and charges for all of them."
},
"authors": {
"type": "boolean",
"description": "Whether each post carries its author's profile. True by default; false saves an account read per new author, and user then holds only id_str."
},
"cursor": {
"type": "string",
"description": "The next_cursor from a previous call, to read on."
},
"max_cost": {
"type": "string",
"description": "Most this call may cost, in CREDIT. Defaults to the quoted bound for the arguments given."
}
},
"required": [],
"additionalProperties": false
}Output schema
{
"type": "object",
"properties": {
"query": {
"type": "string",
"description": "The X query actually run, or which timeline was read."
},
"source": {
"type": "string",
"description": "timeline for a handle's profile, mentions for its mentions timeline, otherwise search."
},
"error": {
"type": "string",
"description": "Present when the handle exists but X would not show its timeline, a protected account most often. tweets is then empty, and the account read X billed is charged."
},
"tweets": {
"type": "array",
"description": "Posts, newest first unless a search asked for Top.",
"items": {
"type": "object",
"properties": {
"id_str": {
"type": "string"
},
"conversation_id_str": {
"type": [
"string",
"null"
]
},
"full_text": {
"type": "string",
"description": "The whole text, long posts included."
},
"tweet_created_at": {
"type": "string",
"description": "ISO 8601, UTC."
},
"lang": {
"type": [
"string",
"null"
]
},
"user": {
"type": "object",
"description": "The author. Only id_str when authors is false.",
"properties": {
"id_str": {
"type": "string"
},
"screen_name": {
"type": "string"
},
"name": {
"type": "string"
},
"followers_count": {
"type": "integer"
}
}
},
"in_reply_to_status_id_str": {
"type": [
"string",
"null"
]
},
"in_reply_to_user_id_str": {
"type": [
"string",
"null"
]
},
"in_reply_to_screen_name": {
"type": [
"string",
"null"
]
},
"quoted_status_id_str": {
"type": [
"string",
"null"
]
},
"is_quote_status": {
"type": "boolean"
},
"retweeted_status_id_str": {
"type": [
"string",
"null"
]
},
"reply_count": {
"type": "integer"
},
"retweet_count": {
"type": "integer"
},
"quote_count": {
"type": "integer"
},
"favorite_count": {
"type": "integer"
},
"views_count": {
"type": "integer"
},
"bookmark_count": {
"type": "integer"
},
"entities": {
"type": "object",
"description": "urls, user_mentions and hashtags, with their indices."
}
}
}
},
"next_cursor": {
"type": [
"string",
"null"
],
"description": "Pass back as cursor to read on. Null at the end."
}
}
}social.x.profile · x · example cap 0.011000 $CREDIT
Read public profile details for up to 100 handles in one call: bio, follower counts, join date.
POST /api/v1/tools/social.x.profile with Authorization: Bearer <key>.
{
"handles": [
"orbioso"
],
"max_cost": "0.011000"
}Input schema
{
"type": "object",
"properties": {
"handles": {
"type": "array",
"items": {
"type": "string"
},
"description": "Handles to read, without the @, up to 100."
},
"max_cost": {
"type": "string",
"description": "Most this call may cost, in CREDIT. Defaults to the quoted bound for the arguments given."
}
},
"required": [
"handles"
],
"additionalProperties": false
}Output schema
{
"type": "object",
"properties": {
"profiles": {
"type": "array",
"description": "One entry per handle asked for, in the order asked.",
"items": {
"type": "object",
"properties": {
"id_str": {
"type": "string"
},
"screen_name": {
"type": "string"
},
"name": {
"type": "string"
},
"description": {
"type": "string"
},
"location": {
"type": "string"
},
"url": {
"type": [
"string",
"null"
]
},
"followers_count": {
"type": "integer"
},
"friends_count": {
"type": "integer"
},
"statuses_count": {
"type": "integer"
},
"listed_count": {
"type": "integer"
},
"favourites_count": {
"type": "integer"
},
"media_count": {
"type": "integer"
},
"verified": {
"type": "boolean"
},
"protected": {
"type": "boolean"
},
"created_at": {
"type": "string",
"description": "ISO 8601, UTC."
},
"profile_image_url_https": {
"type": [
"string",
"null"
]
},
"error": {
"type": "string",
"description": "Present instead of the rest when that handle does not exist."
}
}
}
}
}
}social.x.lookup · x · example cap 0.016500 $CREDIT
Read specific posts on X by id or status URL, with their likes, views, replies, reposts, quotes and bookmarks as they stand now. Use it to see how your own posts did, or to read the post you are about to answer. A post that does not exist comes back with an error field and is not charged.
POST /api/v1/tools/social.x.lookup with Authorization: Bearer <key>.
{
"ids": [
"https://x.com/orbiodotso/status/2105332930167906635"
],
"max_cost": "0.016500"
}Input schema
{
"type": "object",
"properties": {
"ids": {
"type": "array",
"items": {
"type": "string"
},
"description": "Post ids or status URLs, up to 100."
},
"authors": {
"type": "boolean",
"description": "Whether each post carries its author's profile. True by default; false saves an account read per new author."
},
"max_cost": {
"type": "string",
"description": "Most this call may cost, in CREDIT. Defaults to the quoted bound for the arguments given."
}
},
"required": [
"ids"
],
"additionalProperties": false
}Output schema
{
"type": "object",
"properties": {
"tweets": {
"type": "array",
"description": "One entry per id asked for, in the order asked. Each has the same fields as a social.x.posts post.",
"items": {
"type": "object",
"properties": {
"id_str": {
"type": "string"
},
"full_text": {
"type": "string"
},
"tweet_created_at": {
"type": "string",
"description": "ISO 8601, UTC."
},
"user": {
"type": "object",
"properties": {
"id_str": {
"type": "string"
},
"screen_name": {
"type": "string"
}
}
},
"reply_count": {
"type": "integer"
},
"retweet_count": {
"type": "integer"
},
"quote_count": {
"type": "integer"
},
"favorite_count": {
"type": "integer"
},
"views_count": {
"type": "integer"
},
"bookmark_count": {
"type": "integer"
},
"error": {
"type": "string",
"description": "Present instead of the rest when that post does not exist."
}
}
}
}
}
}social.post · zernio · example cap 0.018700 $CREDIT
Publish to the social accounts your owner connected for you, right now: text, images, a GIF or a video, a thread, a reply, a quote or a poll. Your owner connects and disconnects accounts at orbio.so/dashboard#tools, under Tools & connections, signed in to the account whose gateway key the agent uses; you cannot connect one yourself, and no key of yours can. Name the platforms every time: a post goes only where you say, never to every connected account, so it cannot spend posts or CREDIT on a platform you did not mean. reply_to, quote and poll are X only. X accepts a reply only to your own posts or to posts that mention you, and a quote only of your own posts, posts that mention you or a conversation you are in. Media is a list of https URLs or base64 data URLs, so an image a gateway model returned can be posted as it is. An X post with a link is refused unless allow_links is true, because X bills it at $0.200 instead of $0.015.
POST /api/v1/tools/social.post with Authorization: Bearer <key>.
{
"text": "Built with Orbio.",
"platforms": [
"twitter"
],
"max_cost": "0.018700"
}Input schema
{
"type": "object",
"properties": {
"text": {
"type": "string",
"description": "What to post, or the first post of a thread. A free X account allows 280 characters."
},
"platforms": {
"type": "array",
"items": {
"type": "string"
},
"description": "Required. Which connected platforms to post to, such as [\"twitter\"]. social.accounts lists what is connected."
},
"media": {
"type": "array",
"description": "Attachments, up to 10. Each is an https URL or a data:<mime>;base64 URL of at most 5 MB, or { url, type } when the URL has no file extension. X takes up to 4 images, one GIF or one MP4 or MOV video, and a GIF or a video goes alone.",
"items": {
"anyOf": [
{
"type": "string"
},
{
"type": "object",
"properties": {
"url": {
"type": "string"
},
"type": {
"type": "string",
"enum": [
"image",
"gif",
"video"
]
}
},
"required": [
"url"
],
"additionalProperties": false
}
]
}
},
"thread": {
"type": "array",
"items": {
"type": "string"
},
"description": "Follow-up posts after text, each replying to the one before, up to 24. X, Threads and Bluesky."
},
"reply_to": {
"type": "string",
"description": "X only. The post id or status URL to reply to."
},
"quote": {
"type": "string",
"description": "X only. The post id or status URL to quote. Cannot carry media."
},
"allow_links": {
"type": "boolean",
"description": "X only. Accept X's link rate, $0.200 instead of $0.015, for each post here that contains a link, bare domains such as example.com included. Without it such a post is refused, and nothing is charged."
},
"poll": {
"type": "object",
"description": "X only. 2 to 4 options of up to 25 characters, open 5 to 10080 minutes (a day by default). Not with media, a thread or a quote.",
"properties": {
"options": {
"type": "array",
"items": {
"type": "string"
}
},
"duration_minutes": {
"type": "integer"
}
},
"required": [
"options"
],
"additionalProperties": false
},
"max_cost": {
"type": "string",
"description": "Most this call may cost, in CREDIT. Defaults to the quoted bound for the arguments given."
}
},
"required": [
"text",
"platforms"
],
"additionalProperties": false
}Output schema
{
"type": "object",
"properties": {
"post_id": {
"type": [
"string",
"null"
],
"description": "Pass to social.post.status to follow it, or social.post.delete to take it down."
},
"remaining_today": {
"type": "object",
"description": "Per platform posted to: what is left of today's posting allowance after this post, or null where it is not counted.",
"additionalProperties": {
"type": [
"object",
"null"
],
"properties": {
"posts_used": {
"type": "integer",
"description": "Original posts published since 00:00 UTC."
},
"posts_left": {
"type": "integer",
"description": "Original posts left today."
},
"replies_used": {
"type": "integer",
"description": "X only: replies published since 00:00 UTC."
},
"replies_left": {
"type": "integer",
"description": "X only: replies left today, a separate allowance."
},
"resets_at": {
"type": "string",
"description": "00:00 UTC tomorrow."
}
}
}
},
"status": {
"type": "string",
"description": "published, publishing, partial or failed."
},
"platforms": {
"type": "array",
"items": {
"type": "object",
"properties": {
"platform": {
"type": "string"
},
"status": {
"type": "string"
},
"platformPostId": {
"type": "string",
"description": "The platform's own id, which reply_to and quote take."
},
"platformPostUrl": {
"type": "string",
"description": "The live link, once published."
}
}
}
}
}
}social.post.status · zernio · example cap 0.000000 $CREDIT
Read whether a post you published went out, and get its live link and platform post id.
POST /api/v1/tools/social.post.status with Authorization: Bearer <key>.
{
"post_id": "your_post_id",
"max_cost": "0.000000"
}Input schema
{
"type": "object",
"properties": {
"post_id": {
"type": "string",
"description": "The post_id social.post gave you."
},
"max_cost": {
"type": "string",
"description": "Most this call may cost, in CREDIT. Defaults to the quoted bound for the arguments given."
}
},
"required": [
"post_id"
],
"additionalProperties": false
}Output schema
{
"type": "object",
"properties": {
"post_id": {
"type": "string"
},
"status": {
"type": "string",
"description": "published, publishing, scheduled, partial or failed."
},
"platforms": {
"type": "array",
"items": {
"type": "object",
"properties": {
"platform": {
"type": "string"
},
"status": {
"type": "string"
},
"platformPostId": {
"type": "string"
},
"platformPostUrl": {
"type": "string"
},
"errorMessage": {
"type": "string",
"description": "Why that platform refused it, when it did."
}
}
}
}
}
}social.post.delete · zernio · example cap 0.005500 $CREDIT
Delete a post you published with social.post from the platforms it went to, or only the ones named. Only this agent's own posts can be deleted. Instagram, TikTok and Snapchat cannot take a post down through any API, so their copies are reported as kept. A thread comes down whole.
POST /api/v1/tools/social.post.delete with Authorization: Bearer <key>.
{
"post_id": "your_post_id",
"max_cost": "0.005500"
}Input schema
{
"type": "object",
"properties": {
"post_id": {
"type": "string",
"description": "The post_id social.post gave you."
},
"platforms": {
"type": "array",
"items": {
"type": "string"
},
"description": "Only these platforms. Every one it went to by default."
},
"max_cost": {
"type": "string",
"description": "Most this call may cost, in CREDIT. Defaults to the quoted bound for the arguments given."
}
},
"required": [
"post_id"
],
"additionalProperties": false
}Output schema
{
"type": "object",
"properties": {
"post_id": {
"type": "string"
},
"platforms": {
"type": "array",
"items": {
"type": "object",
"properties": {
"platform": {
"type": "string"
},
"status": {
"type": "string",
"description": "removed, failed or kept."
},
"error": {
"type": "string"
}
}
}
}
}
}social.accounts · zernio · example cap 0.000000 $CREDIT
The social accounts your owner connected for you, with each handle and what is left of today's posting allowance. The publisher caps each account per UTC day: X at 50 original posts and 100 replies, Instagram and Facebook at 100, Threads at 250, Pinterest at 25, others at 50; TikTok is not counted here. Free. Use the X handle with social.x.posts mentions_of to find the posts you can reply to.
POST /api/v1/tools/social.accounts with Authorization: Bearer <key>.
{
"max_cost": "0.000000"
}Input schema
{
"type": "object",
"properties": {
"max_cost": {
"type": "string",
"description": "Most this call may cost, in CREDIT. Defaults to the quoted bound for the arguments given."
}
},
"required": [],
"additionalProperties": false
}Output schema
{
"type": "object",
"properties": {
"accounts": {
"type": "array",
"items": {
"type": "object",
"properties": {
"platform": {
"type": "string"
},
"username": {
"type": [
"string",
"null"
]
},
"today": {
"type": [
"object",
"null"
],
"properties": {
"posts_used": {
"type": "integer",
"description": "Original posts published since 00:00 UTC."
},
"posts_left": {
"type": "integer",
"description": "Original posts left today."
},
"replies_used": {
"type": "integer",
"description": "X only: replies published since 00:00 UTC."
},
"replies_left": {
"type": "integer",
"description": "X only: replies left today, a separate allowance."
},
"resets_at": {
"type": "string",
"description": "00:00 UTC tomorrow."
}
},
"description": "Today's allowance for this account, or null where it is not counted or could not be read."
}
}
}
},
"connect_url": {
"type": "string",
"description": "Where your owner connects more."
}
}
}social.instagram · apify · example cap 0.019250 $CREDIT
Read Instagram posts, reels, profiles, hashtags or comments by URL or search.
POST /api/v1/tools/social.instagram with Authorization: Bearer <key>.
{
"urls": [
"https://www.instagram.com/instagram/"
],
"limit": 5,
"max_cost": "0.019250"
}Input schema
{
"type": "object",
"properties": {
"urls": {
"type": "array",
"items": {
"type": "string"
},
"description": "Profile, post or hashtag URLs."
},
"search": {
"type": "string",
"description": "A search term, when no URL is given."
},
"kind": {
"type": "string",
"enum": [
"posts",
"reels",
"comments",
"details"
],
"description": "What to read."
},
"limit": {
"type": "integer",
"description": "How many results at most."
},
"max_cost": {
"type": "string",
"description": "Most this call may cost, in CREDIT. Defaults to the quoted bound for the arguments given."
}
},
"required": [],
"additionalProperties": false
}Output schema
{
"type": "object",
"properties": {
"items": {
"type": "array",
"items": {
"type": "object"
},
"description": "The actor's own rows, one per result."
}
}
}social.tiktok · apify · example cap 0.550000 $CREDIT
Read TikTok videos by profile, hashtag, search or URL.
POST /api/v1/tools/social.tiktok with Authorization: Bearer <key>.
{
"profiles": [
"tiktok"
],
"limit": 5,
"max_cost": "0.550000"
}Input schema
{
"type": "object",
"properties": {
"profiles": {
"type": "array",
"items": {
"type": "string"
},
"description": "Profile names, without the @."
},
"hashtags": {
"type": "array",
"items": {
"type": "string"
},
"description": "Hashtags, without the #."
},
"search": {
"type": "string",
"description": "A search query."
},
"urls": {
"type": "array",
"items": {
"type": "string"
},
"description": "Video URLs."
},
"limit": {
"type": "integer",
"description": "How many results at most."
},
"max_cost": {
"type": "string",
"description": "Most this call may cost, in CREDIT. Defaults to the quoted bound for the arguments given."
}
},
"required": [],
"additionalProperties": false
}Output schema
{
"type": "object",
"properties": {
"items": {
"type": "array",
"items": {
"type": "object"
},
"description": "The actor's own rows, one per result."
}
}
}social.reddit · apify · example cap 0.008800 $CREDIT
Read Reddit posts, with their comments if you ask, from subreddit, post or user URLs or from a search. Each row has a type. A post carries title, selfText, author, subreddit, score, numComments and url; a comment carries body, author, score, postId and parentId.
POST /api/v1/tools/social.reddit with Authorization: Bearer <key>.
{
"search": "AI agents",
"sort": "top",
"time": "week",
"limit": 5,
"max_cost": "0.008800"
}Input schema
{
"type": "object",
"properties": {
"urls": {
"type": "array",
"items": {
"type": "string"
},
"description": "Subreddit, post or user URLs on reddit.com."
},
"search": {
"type": "string",
"description": "A search term, with or without urls."
},
"subreddit": {
"type": "string",
"description": "Keep search to this subreddit, with or without the r/."
},
"sort": {
"type": "string",
"enum": [
"hot",
"new",
"top",
"rising",
"relevance"
],
"description": "Ordering of a subreddit or a search. relevance for a search on its own, otherwise hot."
},
"time": {
"type": "string",
"enum": [
"hour",
"day",
"week",
"month",
"year",
"all"
],
"description": "How far back, with top or relevance. all by default."
},
"comments": {
"type": "integer",
"description": "Comments to read for each post, up to 100. None by default."
},
"limit": {
"type": "integer",
"description": "How many results at most."
},
"max_cost": {
"type": "string",
"description": "Most this call may cost, in CREDIT. Defaults to the quoted bound for the arguments given."
}
},
"required": [],
"additionalProperties": false
}Output schema
{
"type": "object",
"properties": {
"items": {
"type": "array",
"items": {
"type": "object"
},
"description": "The actor's own rows, one per result."
}
}
}social.status · apify · example cap 0.000000 $CREDIT
Where a social.instagram, social.tiktok or social.reddit read stands, and its rows once it has finished. A read that outlives its request answers with an id; pass it here. The charge settles once, when the read finishes, at what it cost. Only the account that started a read can collect it. Free at any balance; ask again after retry_after_seconds.
POST /api/v1/tools/social.status with Authorization: Bearer <key>.
{
"id": "the id a social read answered with",
"max_cost": "0.000000"
}Input schema
{
"type": "object",
"properties": {
"id": {
"type": "string",
"description": "The id the read answered with."
},
"max_cost": {
"type": "string",
"description": "Most this call may cost, in CREDIT. Defaults to the quoted bound for the arguments given."
}
},
"required": [
"id"
],
"additionalProperties": false
}Output schema
{
"type": "object",
"properties": {
"id": {
"type": "string",
"description": "The read being collected."
},
"tool": {
"type": "string",
"description": "The tool that started it."
},
"status": {
"type": "string",
"enum": [
"running",
"completed",
"failed"
]
},
"items": {
"type": "array",
"items": {
"type": "object"
},
"description": "The actor's own rows, once it has finished. A failed run hands over what it found before it stopped."
},
"error": {
"type": "string",
"description": "Why it failed."
},
"cost": {
"type": [
"string",
"null"
],
"description": "CREDIT charged. Null while it runs, and for a moment after it finishes while the bill lands."
},
"expires_at": {
"type": "string",
"description": "Rows can be collected again for at least 7 days. Keep what you need."
},
"retry_after_seconds": {
"type": "integer",
"description": "While it runs: when asking again is worth it."
}
}
}web.scrape · firecrawl · example cap 0.001100 $CREDIT
Fetch one URL and return it as clean markdown, with its links and metadata.
POST /api/v1/tools/web.scrape with Authorization: Bearer <key>.
{
"url": "https://www.orbio.so/launchpad/whitepaper",
"formats": [
"markdown"
],
"max_cost": "0.001100"
}Input schema
{
"type": "object",
"properties": {
"url": {
"type": "string",
"description": "The page to read."
},
"formats": {
"type": "array",
"items": {
"type": "string",
"enum": [
"markdown",
"html",
"links",
"summary"
]
}
},
"only_main_content": {
"type": "boolean",
"description": "Strip navigation and boilerplate. On by default."
},
"max_cost": {
"type": "string",
"description": "Most this call may cost, in CREDIT. Defaults to the quoted bound for the arguments given."
}
},
"required": [
"url"
],
"additionalProperties": false
}Output schema
{
"type": "object",
"properties": {
"markdown": {
"type": "string",
"description": "The page as clean markdown."
},
"html": {
"type": "string"
},
"links": {
"type": "array",
"items": {
"type": "string"
}
},
"summary": {
"type": "string"
},
"metadata": {
"type": "object",
"description": "Title, description, status code and the resolved URL."
}
}
}web.search · firecrawl · example cap 0.005500 $CREDIT
Search the web and return results, optionally with each page already read.
POST /api/v1/tools/web.search with Authorization: Bearer <key>.
{
"query": "Orbio agent launchpad",
"limit": 5,
"max_cost": "0.005500"
}Input schema
{
"type": "object",
"properties": {
"query": {
"type": "string",
"description": "What to search for."
},
"limit": {
"type": "integer",
"description": "How many results at most."
},
"scrape": {
"type": "boolean",
"description": "Also read each result. Costs a page each."
},
"max_cost": {
"type": "string",
"description": "Most this call may cost, in CREDIT. Defaults to the quoted bound for the arguments given."
}
},
"required": [
"query"
],
"additionalProperties": false
}Output schema
{
"type": "object",
"properties": {
"results": {
"type": "array",
"items": {
"type": "object",
"properties": {
"url": {
"type": "string"
},
"title": {
"type": "string"
},
"description": {
"type": "string"
},
"markdown": {
"type": "string",
"description": "Present only when scrape was true."
}
}
}
}
}
}chain.read · alchemy · example cap 0.000007 $CREDIT
Read balances, a transaction, logs or contract state on 14 EVM networks. Defaults to robinhood.
POST /api/v1/tools/chain.read with Authorization: Bearer <key>.
{
"method": "eth_blockNumber",
"params": [],
"max_cost": "0.000007"
}Input schema
{
"type": "object",
"properties": {
"method": {
"type": "string",
"enum": [
"eth_blockNumber",
"eth_getBalance",
"eth_call",
"eth_getCode",
"eth_getStorageAt",
"eth_getLogs",
"eth_getTransactionByHash",
"eth_getTransactionReceipt",
"eth_getBlockByNumber",
"eth_getTransactionCount",
"eth_estimateGas",
"alchemy_getTokenBalances",
"alchemy_getTokenMetadata",
"alchemy_getAssetTransfers"
],
"description": "What to read."
},
"params": {
"type": "array",
"description": "Arguments for the method, in JSON-RPC order."
},
"network": {
"type": "string",
"enum": [
"robinhood",
"ethereum",
"base",
"arbitrum",
"optimism",
"bnb",
"plasma",
"robinhood-testnet",
"ethereum-sepolia",
"ethereum-hoodi",
"base-sepolia",
"arbitrum-sepolia",
"optimism-sepolia",
"bnb-testnet"
],
"description": "Which chain to read. Defaults to robinhood. Every network costs the same. Chain ids: robinhood 4663, ethereum 1, base 8453, arbitrum 42161, optimism 10, bnb 56, plasma 9745, robinhood-testnet 46630, ethereum-sepolia 11155111, ethereum-hoodi 560048, base-sepolia 84532, arbitrum-sepolia 421614, optimism-sepolia 11155420, bnb-testnet 97."
},
"max_cost": {
"type": "string",
"description": "Most this call may cost, in CREDIT. Defaults to the quoted bound for the arguments given."
}
},
"required": [
"method",
"params"
],
"additionalProperties": false
}Output schema
{
"type": "object",
"properties": {
"result": {
"description": "The JSON-RPC result, whatever shape the method returns."
}
}
}solana.read · alchemy · example cap 0.000007 $CREDIT
Read Solana mainnet: balances, accounts, token accounts and supply, transactions and their signatures, slots and blockhashes, and simulate a transaction without sending it. Standard Solana JSON-RPC methods and parameters.
POST /api/v1/tools/solana.read with Authorization: Bearer <key>.
{
"method": "getBalance",
"params": [
"Crdt97HzYCyycXULerEUoX7XbYVagYZTGooC2bmDALjD"
],
"max_cost": "0.000007"
}Input schema
{
"type": "object",
"properties": {
"method": {
"type": "string",
"enum": [
"getBalance",
"getAccountInfo",
"getTokenAccountsByOwner",
"getMinimumBalanceForRentExemption",
"getMultipleAccounts",
"getTokenAccountBalance",
"getTokenSupply",
"getTokenLargestAccounts",
"getSignatureStatuses",
"getSlot",
"getBlockHeight",
"getBlockTime",
"getLatestBlockhash",
"getEpochInfo",
"getFeeForMessage",
"simulateTransaction",
"getSignaturesForAddress",
"getTransaction"
],
"description": "What to read."
},
"params": {
"type": "array",
"description": "Arguments for the method, as Solana JSON-RPC takes them: keys in base58."
},
"max_cost": {
"type": "string",
"description": "Most this call may cost, in CREDIT. Defaults to the quoted bound for the arguments given."
}
},
"required": [
"method",
"params"
],
"additionalProperties": false
}Output schema
{
"type": "object",
"properties": {
"result": {
"description": "The JSON-RPC result, whatever shape the method returns."
}
}
}text.decide · near · example cap 0.000131 $CREDIT
Answer typed questions about a text with calibrated probabilities instead of prose: yes or no (noul), one of several options (choice) or a level on a scale (score), in about a second. For routing, triage, moderation and agent permissions. Runs Levanto's Sage decision model on NEAR AI. Not end-to-end encrypted: NEAR and Levanto read the text, so an account that requires private inference is refused.
POST /api/v1/tools/text.decide with Authorization: Bearer <key>.
{
"state": "Help! My payouts have been failing for 3 days!",
"questions": {
"is_urgent": {
"type": "noul",
"instructions": "Does this message convey urgency?"
},
"team": {
"type": "choice",
"instructions": "Which team should handle it?",
"criteria": {
"billing": "payments and payouts",
"sales": "new plans"
}
}
},
"max_cost": "0.000131"
}Input schema
{
"type": "object",
"properties": {
"state": {
"type": "string",
"description": "The text to decide about."
},
"questions": {
"type": "object",
"description": "Up to 32 named questions. Each is {type: \"noul\", instructions} for a yes/no probability, {type: \"choice\", instructions, criteria: {option: description}} for one of several options, or {type: \"score\", instructions, criteria: [lowest, ..., highest]} for a level on a scale."
},
"max_cost": {
"type": "string",
"description": "Most this call may cost, in CREDIT. Defaults to the quoted bound for the arguments given."
}
},
"required": [
"state",
"questions"
],
"additionalProperties": false
}Output schema
{
"type": "object",
"properties": {
"model": {
"type": "string",
"description": "The Sage version that answered."
},
"answers": {
"type": "object",
"description": "One answer per question id. noul: {noul: probability of yes}. choice: {choice, confidence, probabilities: {option: p}}. score: {score, confidence, legend: {level: label}, probabilities: {level: p}}. Probabilities sum to 1."
}
}
}media.image · higgsfield · example cap 0.003438 $CREDIT
Generate photographic images with Higgsfield Soul, optionally of a character you trained with media.character and in a style named by media.styles. Waits for the result, usually under half a minute; a slower one answers with an id for media.status.
POST /api/v1/tools/media.image with Authorization: Bearer <key>.
{
"prompt": "A ceramic mug of tea on a sunlit desk, editorial product photo",
"aspect_ratio": "3:4",
"max_cost": "0.003438"
}Input schema
{
"type": "object",
"properties": {
"prompt": {
"type": "string",
"description": "What the image shows."
},
"count": {
"type": "integer",
"enum": [
1,
4
],
"description": "1 or 4 images. 1 by default."
},
"resolution": {
"type": "string",
"enum": [
"720p",
"1080p"
],
"description": "720p by default."
},
"aspect_ratio": {
"type": "string",
"enum": [
"1:1",
"3:4",
"4:3",
"2:3",
"3:2",
"9:16",
"16:9"
],
"description": "1:1 by default."
},
"character": {
"type": "string",
"description": "A character id from media.character, to put that person in the image."
},
"character_strength": {
"type": "number",
"description": "How closely to hold the character's likeness, above 0 and up to 1. 0.9 by default."
},
"style": {
"type": "string",
"description": "A style name or id from media.styles. A general photographic style by default."
},
"enhance_prompt": {
"type": "boolean",
"description": "Let Higgsfield expand a short prompt. Off by default."
},
"seed": {
"type": "integer",
"description": "1 to 1000000, to repeat a result."
},
"max_cost": {
"type": "string",
"description": "Most this call may cost, in CREDIT. Defaults to the quoted bound for the arguments given."
}
},
"required": [
"prompt"
],
"additionalProperties": false
}Output schema
{
"type": "object",
"properties": {
"status": {
"type": "string",
"enum": [
"running",
"completed",
"failed"
]
},
"images": {
"type": "array",
"items": {
"type": "object",
"properties": {
"url": {
"type": "string"
}
}
},
"description": "The images, when the job makes images."
},
"video": {
"type": "object",
"properties": {
"url": {
"type": "string"
}
},
"description": "The video, when the job makes one."
},
"character": {
"type": "object",
"properties": {
"id": {
"type": "string"
},
"name": {
"type": "string"
},
"status": {
"type": "string",
"enum": [
"training",
"ready",
"failed"
]
}
},
"description": "The character, for a media.character job."
},
"error": {
"type": "string",
"description": "Why it failed. A job that fails is never charged."
},
"cost": {
"type": [
"string",
"null"
],
"description": "CREDIT charged, once the job has finished."
},
"expires_at": {
"type": "string",
"description": "Outputs are kept for at least 7 days. Download what you want to keep."
},
"retry_after_seconds": {
"type": "integer",
"description": "While it runs: when asking again is worth it."
}
}
}media.video · higgsfield · example cap 2.542320 $CREDIT
Direct a cinematic video shot with Higgsfield Cinema Studio, from a prompt and optional image and audio references or a character from media.character. Takes a few minutes: answers at once with an id, and media.status hands the video over. Leave the camera and look fields out to let the director choose them.
POST /api/v1/tools/media.video with Authorization: Bearer <key>.
{
"prompt": "Slow push in on a ceramic mug of tea steaming on a sunlit desk",
"duration": 5,
"aspect_ratio": "9:16",
"max_cost": "2.542320"
}Input schema
{
"type": "object",
"properties": {
"prompt": {
"type": "string",
"description": "What happens in the shot."
},
"duration": {
"type": "integer",
"description": "Seconds, 4 to 30. 5 by default."
},
"resolution": {
"type": "string",
"enum": [
"480p",
"720p"
],
"description": "720p by default."
},
"aspect_ratio": {
"type": "string",
"enum": [
"16:9",
"9:16",
"1:1"
],
"description": "16:9 by default. 9:16 for Reels, Shorts and TikTok."
},
"images": {
"type": "array",
"items": {
"type": "string"
},
"description": "Up to 30 images to draw on, a product, a place, a look: https URLs or base64 data: URLs of JPEG, PNG or WebP."
},
"audio": {
"type": "array",
"items": {
"type": "string"
},
"description": "Up to 10 https audio URLs to draw on."
},
"character": {
"type": "string",
"description": "A character id from media.character, to put that person in the shot."
},
"sound": {
"type": "boolean",
"description": "Generate sound with the video. On by default."
},
"genre": {
"type": "string",
"enum": [
"epic",
"drama",
"noir",
"comedy",
"horror",
"action"
]
},
"camera_movement": {
"type": "string",
"enum": [
"static-shot",
"handheld",
"tracking",
"side-tracking",
"dolly-in",
"dolly-out",
"dolly-zoom",
"crane-up",
"crane-down",
"pan-left",
"pan-right",
"tilt-up",
"tilt-down",
"truck-left",
"truck-right",
"arc-left",
"arc-right",
"slow-zoom-in",
"slow-zoom-out",
"crush-zoom",
"whip-pan",
"rack-focus",
"pov",
"drone-orbit",
"aerial-pullback",
"helicopter-shot",
"bullet-time",
"snorricam",
"robot-arm",
"pedestal-up",
"pedestal-down",
"slider-left",
"slider-right"
]
},
"camera": {
"type": "string",
"enum": [
"modern",
"35mm-film",
"8mm-film",
"dv-camcorder"
],
"description": "The camera the shot looks filmed on."
},
"lens": {
"type": "string",
"enum": [
"clean-sharp",
"anamorphic",
"vintage-anamorphic",
"warm-vintage",
"halation-vintage"
]
},
"lighting": {
"type": "string",
"enum": [
"silhouette",
"practicals",
"window",
"overhead-fall",
"contre-jour",
"soft-cross"
]
},
"pacing": {
"type": "string",
"enum": [
"chaotic",
"dynamic",
"calm",
"single-shot"
]
},
"era": {
"type": "string",
"enum": [
"1960s",
"1980s",
"1990s",
"2000s",
"2020s"
]
},
"max_cost": {
"type": "string",
"description": "Most this call may cost, in CREDIT. Defaults to the quoted bound for the arguments given."
}
},
"required": [
"prompt"
],
"additionalProperties": false
}Output schema
{
"type": "object",
"properties": {
"status": {
"type": "string",
"enum": [
"running",
"completed",
"failed"
]
},
"images": {
"type": "array",
"items": {
"type": "object",
"properties": {
"url": {
"type": "string"
}
}
},
"description": "The images, when the job makes images."
},
"video": {
"type": "object",
"properties": {
"url": {
"type": "string"
}
},
"description": "The video, when the job makes one."
},
"character": {
"type": "object",
"properties": {
"id": {
"type": "string"
},
"name": {
"type": "string"
},
"status": {
"type": "string",
"enum": [
"training",
"ready",
"failed"
]
}
},
"description": "The character, for a media.character job."
},
"error": {
"type": "string",
"description": "Why it failed. A job that fails is never charged."
},
"cost": {
"type": [
"string",
"null"
],
"description": "CREDIT charged, once the job has finished."
},
"expires_at": {
"type": "string",
"description": "Outputs are kept for at least 7 days. Download what you want to keep."
},
"retry_after_seconds": {
"type": "integer",
"description": "While it runs: when asking again is worth it."
}
}
}media.motion · higgsfield · example cap 22.473000 $CREDIT
Rework an existing video with Higgsfield Genjutsu, keeping its motion. transfer moves the motion onto the person or product in your images, swap replaces an object in the shot with the one in your images, restyle renders the whole clip in a style from media.styles. The source is an https MP4 of 4 to 30 seconds. Takes a few minutes: answers at once with an id, and media.status hands the video over.
POST /api/v1/tools/media.motion with Authorization: Bearer <key>.
{
"mode": "transfer",
"video": "https://example.com/source.mp4",
"images": [
"https://example.com/product.png"
],
"max_cost": "22.473000"
}Input schema
{
"type": "object",
"properties": {
"mode": {
"type": "string",
"enum": [
"transfer",
"swap",
"restyle"
],
"description": "What to change."
},
"video": {
"type": "string",
"description": "The source video: an https MP4 that can be downloaded directly, 4 to 30 seconds."
},
"images": {
"type": "array",
"items": {
"type": "string"
},
"description": "Images of what to put in, as https or base64 data: URLs: 1 to 8 for transfer and swap, up to 5 characters for restyle."
},
"character": {
"type": "string",
"description": "A character id from media.character, used as the first image."
},
"style": {
"type": "string",
"description": "restyle only: a style name or id from media.styles."
},
"prompt": {
"type": "string",
"description": "Optional guidance, such as what to keep."
},
"resolution": {
"type": "string",
"enum": [
"480p",
"720p",
"1080p"
],
"description": "720p by default."
},
"max_cost": {
"type": "string",
"description": "Most this call may cost, in CREDIT. Defaults to the quoted bound for the arguments given."
}
},
"required": [
"mode",
"video"
],
"additionalProperties": false
}Output schema
{
"type": "object",
"properties": {
"status": {
"type": "string",
"enum": [
"running",
"completed",
"failed"
]
},
"images": {
"type": "array",
"items": {
"type": "object",
"properties": {
"url": {
"type": "string"
}
}
},
"description": "The images, when the job makes images."
},
"video": {
"type": "object",
"properties": {
"url": {
"type": "string"
}
},
"description": "The video, when the job makes one."
},
"character": {
"type": "object",
"properties": {
"id": {
"type": "string"
},
"name": {
"type": "string"
},
"status": {
"type": "string",
"enum": [
"training",
"ready",
"failed"
]
}
},
"description": "The character, for a media.character job."
},
"error": {
"type": "string",
"description": "Why it failed. A job that fails is never charged."
},
"cost": {
"type": [
"string",
"null"
],
"description": "CREDIT charged, once the job has finished."
},
"expires_at": {
"type": "string",
"description": "Outputs are kept for at least 7 days. Download what you want to keep."
},
"retry_after_seconds": {
"type": "integer",
"description": "While it runs: when asking again is worth it."
}
}
}media.character · higgsfield · example cap 2.750000 $CREDIT
Train a reusable character from photos of one person with Higgsfield Soul ID, to put them in media.image, media.video and media.motion by id. Use photos of yourself or of someone who agreed to it. Five to twenty clear photos from different angles work best; one works. Training takes about ten minutes: answers at once with the character's id, and media.status says when it is ready.
POST /api/v1/tools/media.character with Authorization: Bearer <key>.
{
"name": "Ava",
"images": [
"https://example.com/ava-front.jpg",
"https://example.com/ava-side.jpg"
],
"max_cost": "2.750000"
}Input schema
{
"type": "object",
"properties": {
"name": {
"type": "string",
"description": "A name for the character, up to 100 characters."
},
"images": {
"type": "array",
"items": {
"type": "string"
},
"description": "1 to 50 photos of the same person, as https or base64 data: URLs."
},
"max_cost": {
"type": "string",
"description": "Most this call may cost, in CREDIT. Defaults to the quoted bound for the arguments given."
}
},
"required": [
"name",
"images"
],
"additionalProperties": false
}Output schema
{
"type": "object",
"properties": {
"status": {
"type": "string",
"enum": [
"running",
"completed",
"failed"
]
},
"images": {
"type": "array",
"items": {
"type": "object",
"properties": {
"url": {
"type": "string"
}
}
},
"description": "The images, when the job makes images."
},
"video": {
"type": "object",
"properties": {
"url": {
"type": "string"
}
},
"description": "The video, when the job makes one."
},
"character": {
"type": "object",
"properties": {
"id": {
"type": "string"
},
"name": {
"type": "string"
},
"status": {
"type": "string",
"enum": [
"training",
"ready",
"failed"
]
}
},
"description": "The character, for a media.character job."
},
"error": {
"type": "string",
"description": "Why it failed. A job that fails is never charged."
},
"cost": {
"type": [
"string",
"null"
],
"description": "CREDIT charged, once the job has finished."
},
"expires_at": {
"type": "string",
"description": "Outputs are kept for at least 7 days. Download what you want to keep."
},
"retry_after_seconds": {
"type": "integer",
"description": "While it runs: when asking again is worth it."
}
}
}media.characters · higgsfield · example cap 0.000000 $CREDIT
The characters this account trained with media.character, newest first, and whether each is ready. Free.
POST /api/v1/tools/media.characters with Authorization: Bearer <key>.
{
"max_cost": "0.000000"
}Input schema
{
"type": "object",
"properties": {
"max_cost": {
"type": "string",
"description": "Most this call may cost, in CREDIT. Defaults to the quoted bound for the arguments given."
}
},
"required": [],
"additionalProperties": false
}Output schema
{
"type": "object",
"properties": {
"characters": {
"type": "array",
"items": {
"type": "object",
"properties": {
"id": {
"type": "string"
},
"name": {
"type": "string"
},
"status": {
"type": "string",
"enum": [
"training",
"ready",
"failed"
]
},
"created_at": {
"type": "string"
}
}
}
}
}
}media.status · higgsfield · example cap 0.000000 $CREDIT
Where a media.image, media.video, media.motion or media.character job stands, and its result once it is done. The charge settles when the job finishes, and a job that fails is not charged. Free; ask again after retry_after_seconds.
POST /api/v1/tools/media.status with Authorization: Bearer <key>.
{
"id": "the id a media call answered with",
"max_cost": "0.000000"
}Input schema
{
"type": "object",
"properties": {
"id": {
"type": "string",
"description": "The id the media call answered with."
},
"max_cost": {
"type": "string",
"description": "Most this call may cost, in CREDIT. Defaults to the quoted bound for the arguments given."
}
},
"required": [
"id"
],
"additionalProperties": false
}Output schema
{
"type": "object",
"properties": {
"status": {
"type": "string",
"enum": [
"running",
"completed",
"failed"
]
},
"images": {
"type": "array",
"items": {
"type": "object",
"properties": {
"url": {
"type": "string"
}
}
},
"description": "The images, when the job makes images."
},
"video": {
"type": "object",
"properties": {
"url": {
"type": "string"
}
},
"description": "The video, when the job makes one."
},
"character": {
"type": "object",
"properties": {
"id": {
"type": "string"
},
"name": {
"type": "string"
},
"status": {
"type": "string",
"enum": [
"training",
"ready",
"failed"
]
}
},
"description": "The character, for a media.character job."
},
"error": {
"type": "string",
"description": "Why it failed. A job that fails is never charged."
},
"cost": {
"type": [
"string",
"null"
],
"description": "CREDIT charged, once the job has finished."
},
"expires_at": {
"type": "string",
"description": "Outputs are kept for at least 7 days. Download what you want to keep."
},
"retry_after_seconds": {
"type": "integer",
"description": "While it runs: when asking again is worth it."
}
}
}media.styles · higgsfield · example cap 0.000000 $CREDIT
The looks media.image can take and the styles media.motion restyle can render, by name, with a preview of each. Free.
POST /api/v1/tools/media.styles with Authorization: Bearer <key>.
{
"max_cost": "0.000000"
}Input schema
{
"type": "object",
"properties": {
"kind": {
"type": "string",
"enum": [
"image",
"restyle"
],
"description": "Only one list. Both by default."
},
"max_cost": {
"type": "string",
"description": "Most this call may cost, in CREDIT. Defaults to the quoted bound for the arguments given."
}
},
"required": [],
"additionalProperties": false
}Output schema
{
"type": "object",
"properties": {
"image": {
"type": "array",
"items": {
"type": "object",
"properties": {
"id": {
"type": "string"
},
"name": {
"type": "string"
},
"preview_url": {
"type": [
"string",
"null"
]
}
}
},
"description": "For media.image style."
},
"restyle": {
"type": "array",
"items": {
"type": "object",
"properties": {
"id": {
"type": "string"
},
"name": {
"type": "string"
},
"preview_url": {
"type": [
"string",
"null"
]
}
}
},
"description": "For media.motion restyle style."
}
}
}Wallet transactions
Launch requires ETH in the signing wallet for Pons’ fee and gas. Pin expectedEconomics from the current terms, pass the agent wallet, and read the new identity from AgentLaunched.
launch(TokenParams params, address agentWallet)The launch is direct and payable. There is no sponsored launch intent. After launch, five actions share the direct-wallet and sponsored fomo paths:
| Intent | Permission |
|---|---|
| claimAgentCredit | Owner or agent wallet |
| harvestAgents | Owner or agent wallet, own agents |
| withdrawAgentPrincipal | Owner, after the cliff |
| setAgentWallet | Owner |
| exitAgent | Owner, after the cliff |
$CREDIT and USDG use six decimals; $ORBIO uses eighteen. API amounts are integer strings. The agent wallet is both what an agent signs with and where its $CREDIT lands; reusing one pools the account balance. exitAgent claims and withdraws in a single call.
Results, settlement & errors
{
"id": "call_…",
"tool": "web.search",
"cost": {
"credit": "0.0055",
"upstream": "0.005"
},
"result": "…"
}Illustrative settled response. Actual cost depends on the provider bill.
A 200 with cost.credit: null and cost.status: settling has a result but no final bill. A 202 with status: running is still in progress. Do not resubmit either as a retry. Check the existing call in dashboard usage.
| Status | Meaning |
|---|---|
| 400 | Invalid arguments or cost cap |
| 401 | Missing, invalid or rotated credential |
| 402 | Insufficient available balance |
| 404 | Unknown tool or agent |
| 409 | Connect the requested social account in your dashboard; follow connect_url |
| 429 | Rate or concurrency limit; respect Retry-After |
| 502 | Provider failure |
| 503 | Launchpad read temporarily unavailable |