curl --request PATCH \
--url https://developer.creao.ai/v1/agents/{agent_id} \
--header 'Authorization: Bearer <token>' \
--header 'Content-Type: application/json' \
--data '
{
"default_model_id": "anthropic/claude-sonnet-5-5",
"default_thinking_level": "high"
}
'const url = 'https://developer.creao.ai/v1/agents/{agent_id}';
const options = {
method: 'PATCH',
headers: {Authorization: 'Bearer <token>', 'Content-Type': 'application/json'},
body: JSON.stringify({
default_model_id: 'anthropic/claude-sonnet-5-5',
default_thinking_level: 'high'
})
};
fetch(url, options)
.then(res => res.json())
.then(json => console.log(json))
.catch(err => console.error(err));import requests
url = "https://developer.creao.ai/v1/agents/{agent_id}"
payload = {
"default_model_id": "anthropic/claude-sonnet-5-5",
"default_thinking_level": "high"
}
headers = {
"Authorization": "Bearer <token>",
"Content-Type": "application/json"
}
response = requests.patch(url, json=payload, headers=headers)
print(response.text)Update agent settings
Change a personal CREAO agent’s write autonomy and default model.
curl --request PATCH \
--url https://developer.creao.ai/v1/agents/{agent_id} \
--header 'Authorization: Bearer <token>' \
--header 'Content-Type: application/json' \
--data '
{
"default_model_id": "anthropic/claude-sonnet-5-5",
"default_thinking_level": "high"
}
'const url = 'https://developer.creao.ai/v1/agents/{agent_id}';
const options = {
method: 'PATCH',
headers: {Authorization: 'Bearer <token>', 'Content-Type': 'application/json'},
body: JSON.stringify({
default_model_id: 'anthropic/claude-sonnet-5-5',
default_thinking_level: 'high'
})
};
fetch(url, options)
.then(res => res.json())
.then(json => console.log(json))
.catch(err => console.error(err));import requests
url = "https://developer.creao.ai/v1/agents/{agent_id}"
payload = {
"default_model_id": "anthropic/claude-sonnet-5-5",
"default_thinking_level": "high"
}
headers = {
"Authorization": "Bearer <token>",
"Content-Type": "application/json"
}
response = requests.patch(url, json=payload, headers=headers)
print(response.text)Authorizations
Body
External write autonomy for guarded connector actions (write and delete actions in connected services). full_auto performs them immediately. A run started through /v1 cannot collect an approval, so on those runs approval_required and blocked both leave the action unperformed; the run still completes and the agent is told why. In the browser app, approval_required waits for your approval. Reads always run.
full_auto, approval_required, blocked Model the agent runs with when a run omits model (for example, anthropic/claude-sonnet-5-5). Accepts any model offered in the CREAO app's model picker, including models not listed on https://agent.docs.creao.ai/developer-api/supported-models. Setting it pins this exact model, replacing automatic model selection. Changing the model without also sending default_thinking_level clears the saved thinking level. Unknown ids return 400 MODEL_NOT_SUPPORTED; models not available on your plan return 403 MODEL_NOT_ALLOWED.
1 - 255Thinking level for the agent's default model. null clears it so the model's own default applies. Requires a pinned model: when the agent's model is chosen automatically, send default_model_id in the same request or the update returns 400 AGENT_THINKING_NOT_APPLICABLE.
low, medium, high, xhigh, max Response
Agent updated.
External write autonomy for guarded connector actions (write and delete actions in connected services). full_auto performs them immediately. A run started through /v1 cannot collect an approval, so on those runs approval_required and blocked both leave the action unperformed; the run still completes and the agent is told why. In the browser app, approval_required waits for your approval. Reads always run.
full_auto, approval_required, blocked