Octagon Podcasts Agent

Model Name: octagon-podcasts-agent

Answers questions from what investors, strategists, operators and guests said on finance and business podcasts, citing the episode behind every claim.

Capabilities

• 70+ finance and business shows, including Bloomberg Surveillance, All-In, Odd Lots, Invest Like the Best, Excess Returns and Thoughts on the Market
• Episodes from July 2025 onward
• Multi-perspective answers: where views agree, where they disagree, and the reasoning behind each
• Time windows, shows, hosts, guests, companies and tickers read from the question
• Mention counts for "most discussed" questions
• Episode-level citations: show, episode title, date and link, deep-linked to the chapter when the episode has chapters

Use Cases

The Podcasts Agent is best for:

• Catching up on what investors discussed in a given period
• Mapping who is bullish or bearish on a macro topic, and why
• Tracking commentary on a company or ticker across shows
• Following what a named strategist or investor said as a guest
• Summarizing the latest episodes of a show
• Seeing how sentiment on a topic shifted over time

Example Queries

• "What are podcast guests saying about NVDA since August?"
• "Will the Fed keep hiking? Who disagrees?"
• "Summarize the last 3 All-In episodes on AI capex."
• "What has David Rosenberg said as a guest since September?"
• "Which companies were most discussed this week?"
• "How has sentiment on 24/7 trading shifted since September?"

Code Examples

Python

from openai import OpenAI

client = OpenAI(
    api_key="your-octagon-api-key",
    base_url="https://api.octagonai.co/v1"
)

response = client.responses.create(
    model="octagon-podcasts-agent",
    input="What are podcast guests saying about 24/7 trading? Last 60 days."
)

print(response.output[0].content[0].text)
for annotation in response.output[0].content[0].annotations:
    print(annotation.order, annotation.name, annotation.url)

JavaScript

import OpenAI from 'openai';

const client = new OpenAI({
  apiKey: 'your-octagon-api-key',
  baseURL: 'https://api.octagonai.co/v1'
});

const response = await client.responses.create({
  model: "octagon-podcasts-agent",
  input: "What are podcast guests saying about 24/7 trading? Last 60 days."
});

console.log(response.output[0].content[0].text);
for (const annotation of response.output[0].content[0].annotations) {
  console.log(annotation.order, annotation.name, annotation.url);
}

cURL

curl -X POST https://api.octagonai.co/v1/responses \
  -H "Content-Type: application/json" \
  -H "Authorization: Bearer your-octagon-api-key" \
  -d '{
    "model": "octagon-podcasts-agent",
    "input": "What are podcast guests saying about 24/7 trading? Last 60 days.",
    "stream": true
  }' \
  --no-buffer

Citations

Each [n] marker in the answer matches one annotation, and the answer ends with the same list under Sources:.

Field	Description
order	Citation number, matching the [n] marker in the text.
name	Podcast | Episode title | YYYY-MM-DD
url	Episode URL, with &t=<seconds> pointing at the chapter when the episode has chapters.

Example:

1. The Dentist Money Show | Two Cents 9/26 - Trading 24/7? | 2026-09-26: https://www.youtube.com/watch?v=pqZg1dYX7VQ&t=173

How to Create Effective Prompts for Podcast Research

1. Name the time window: Without one, the agent searches the last 30 days.
   • Example: "What did guests say about private credit since July?"
2. Name the show, host or guest to focus on specific voices.
   • Example: "What has Michael Hartnett said on podcasts since August?"
3. Use company names or tickers for company questions.
   • Example: "What are podcast guests saying about Nvidia since Q3 earnings?"
4. Ask for the last N episodes to summarize a show.
   • Example: "Summarize the last 2 Odd Lots episodes."