ant:searchAnswers grounded in live web results, with citations.
ant:search runs a web search on the question before the model sees it and injects the results into context, so time-sensitive answers come from fresh sources rather than training data. The search backend is a priority-ordered chain that fails over between providers, and the web_search tool stays available for follow-up queries. Equivalent to passing tools: [{ type: "ant:search" }] against any model.
Live routing over the last 30 days
Requests
1
Success rate
0%
Avg latency
893ms
23,776 models across 34 providers currently qualify. Per-request limits (context size, cost caps, your model allowlist) narrow this further at routing time.
o1-pro
o1-pro
o3
Claude Sonnet 5
o3
OpenAI: o4 Mini High
Anthropic: Claude Opus 4.1 (batch)
Anthropic: Claude Opus 4.6 (batch)
Anthropic: Claude Opus 4.8 (batch)
o1
Anthropic: Claude Opus 4.1
Anthropic: Claude Opus 4.5
GPT-5
OpenAI: o3 Mini High
Anthropic: Claude Opus 5
Anthropic: Claude Opus Latest
OpenAI: o4 Mini (batch)
Anthropic: Claude Opus 4.7
Anthropic: Claude Opus 4.6
Anthropic: Claude Opus 5.5
Anthropic: Claude Opus 5.5 (batch)
Anthropic: Claude Opus 4.8
Anthropic: Claude Opus 4.7 (batch)
Anthropic: Claude Opus 4.5 (batch)
Showing the 24 highest-scoring of 23,776.
from openai import OpenAI
client = OpenAI(
base_url="https://api.antbase.ai/v1",
api_key="YOUR_ANT_API_KEY",
)
response = client.chat.completions.create(
model="ant:search",
messages=[{"role": "user", "content": "Hello!"}]
)
print(response.choices[0].message.content)Every virtual pool accepts the same request shape — swap the model id to change the routing policy.
Browse all virtual models →