> ## Documentation Index
> Fetch the complete documentation index at: https://developers.t2000.ai/llms.txt
> Use this file to discover all available pages before exploring further.

# What's new in AI this week

> A weekly digest of the AI research that actually matters — specific papers with arXiv IDs, plus the debates and launches to track. Exa + Perplexity find it, Claude distills it. Paid as you go.

Keep up without doom-scrolling. Your agent pulls the week's notable AI papers (with arXiv IDs, dates, authors), surfaces the launches and debates worth tracking, and hands back a tight digest with a one-line "why it matters" per item. Self-contained — no inputs needed. \~\$0.06.

<Note>
  A **durable gateway demo**: this is live, web-grounded retrieval a code sandbox can't fabricate. The value is paid semantic search + real-time citations, not generation — exactly the kind of recipe that earns its keep regardless of how capable the client is.
</Note>

***

## The prompt

```
Use t2 services. Using Exa and Perplexity, pull the most important new AI research from the
past week — arXiv papers, lab blogs, notable launches. Give me 5-8 specific
papers with arXiv IDs, dates, and authors, plus 3 debates or launches to
track. One line on why each matters.
```

***

## What runs

1. `POST /exa/v1/search` — semantic search for recent papers + primary pages (\~\$0.02)
2. `POST /perplexity/v1/chat/completions` — web-grounded "what's notable this week" with citations (\~\$0.02)
3. `POST /anthropic/v1/messages` — Claude distills the digest, keeping IDs + sources (\~\$0.02)

***

## Run it

### Claude Desktop (MCP)

```bash theme={null}
npm install -g @t2000/cli && t2 init && t2 fund && t2 mcp install
```

Paste the prompt. Narrow it any way you like — *"...focused on agents and tool use"* or *"...just inference + serving"*.

### SDK

```typescript theme={null}
import { T2000 } from '@t2000/sdk';

const agent = await T2000.create();
const focus = 'LLM agents, reasoning, and inference efficiency';

const [papers, landscape] = await Promise.all([
  agent.pay({
    url: 'https://mpp.t2000.ai/exa/v1/search',
    method: 'POST',
    body: JSON.stringify({
      query: `notable AI research papers past week ${focus} arxiv`,
      numResults: 10,
      contents: { text: true },
    }),
  }),
  agent.pay({
    url: 'https://mpp.t2000.ai/perplexity/v1/chat/completions',
    method: 'POST',
    body: JSON.stringify({
      model: 'sonar',
      messages: [{
        role: 'user',
        content: `What are the most important AI developments and debates this past week (${focus})? Cite sources.`,
      }],
    }),
  }),
]);

const digest = await agent.pay({
  url: 'https://mpp.t2000.ai/anthropic/v1/messages',
  method: 'POST',
  headers: { 'anthropic-version': '2023-06-01' },
  body: JSON.stringify({
    model: 'claude-sonnet-4-5',
    max_tokens: 1200,
    messages: [{
      role: 'user',
      content:
        `Write a weekly AI digest. List 5-8 specific papers with arXiv ID, date, and authors ` +
        `(only ones you can tie to a source — no invented IDs), then 3 debates/launches to track. ` +
        `One line on why each matters.\n\n` +
        `PAPERS: ${JSON.stringify(papers.body)}\n\nLANDSCAPE: ${JSON.stringify(landscape.body)}`,
    }],
  }),
});

console.log((digest.body as { content: { text: string }[] }).content[0].text);
```

***

## Expected output

```
2-3 calls · ~$0.06 · ~8s · 0 taps
5-8 papers w/ arXiv IDs + 3 things to track, each with why-it-matters
```

***

## Extend it

* Add **Jina Reader** (`/jina/v1/read`) to pull the raw arXiv "recent" listing as clean markdown and ground the IDs verbatim
* Use **Firecrawl** (`/firecrawl/v1/scrape`) to read a specific lab blog (e.g. a new model card) in full
* Pipe the digest into **Resend** (`/resend/v1/emails`) for a Monday-morning email to yourself, scheduled via cron
* Chain the **Research with citations** recipe on any single paper to go deeper with sources