> ## Documentation Index
> Fetch the complete documentation index at: https://docs.loom.aayanmishra.com/llms.txt
> Use this file to discover all available pages before exploring further.

# Models — Apollo & Daedalus

> Model cards and guidance for the Apollo family and Daedalus family (code models).

# Apollo model family

Loom Labs provides the Apollo family of models tuned for different latency, cost, and capability trade-offs, as well as the Daedalus family optimized for code generation and reasoning. All Apollo models support a 128k token context window and are available in the Loom AI API as `Apollo-1-2B`, `Apollo-1-4B`, and `Apollo-1-8B`.

The Daedalus family (`Daedalus-1-2B`, `Daedalus-1-8B`) is purpose-built for code reasoning, program synthesis, and debugging workflows.

<CardGroup columns={3}>
  <Card title="Apollo-1 2B" subtitle="Model ID: Apollo-1-2B" description="Lightweight model for low-cost inference and simple tasks. 128k context." />

  <Card title="Apollo-1 4B" subtitle="Model ID: Apollo-1-4B" description="Balanced model for production use-cases. 128k context." />

  <Card title="Apollo-1 8B" subtitle="Model ID: Apollo-1-8B" description="High-capacity model for advanced tasks and deep reasoning. 128k context." />
</CardGroup>

<CardGroup columns={2}>
  <Card title="Daedalus-1 2B" subtitle="Model ID: Daedalus-1-2B" description="2B parameter code reasoning model optimized for structured code generation and debugging." />

  <Card title="Daedalus-1 8B" subtitle="Model ID: Daedalus-1-8B" description="8B parameter code model for advanced code synthesis and reasoning with improved safety alignment." />
</CardGroup>

## Apollo-1 2B

* Model ID: `Apollo-1-2B`
* Availability: Free (Beta)
* Context window: 128k tokens
* Recommended use-cases: classification, short summaries, simple chatbots, low-cost inference

### Example

```bash theme={null}
curl https://api.loom.aayanmishra.com/api/v1/responses \
  -H "Authorization: Bearer $LOOM_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"model":"Apollo-1-2B","input":"Classify the sentiment: I love this product"}'
```

## Apollo-1 4B

* Model ID: `Apollo-1-4B`
* Availability: Free (Beta)
* Context window: 128k tokens
* Recommended use-cases: content generation, multi-turn dialogue, question answering

### Example

```python theme={null}
import requests, os
BASE = "https://api.loom.aayanmishra.com/api/v1"
KEY = os.environ['LOOM_API_KEY']
resp = requests.post(f"{BASE}/chat/completions", headers={"Authorization": f"Bearer {KEY}", "Content-Type": "application/json"}, json={"model":"Apollo-1-4B","messages":[{"role":"user","content":"Explain quantum entanglement in simple terms."}]})
print(resp.json())
```

## Apollo-1 8B

* Model ID: `Apollo-1-8B`
* Availability: Free (Beta)
* Context window: 128k tokens
* Recommended use-cases: advanced content generation, deep reasoning, long-form analysis

### Example

```js theme={null}
// Node example
const res = await fetch('https://api.loom.aayanmishra.com/api/v1/responses', {
  method: 'POST',
  headers: { 'Authorization': `Bearer ${process.env.LOOM_API_KEY}`, 'Content-Type': 'application/json' },
  body: JSON.stringify({ model: 'Apollo-1-8B', input: 'Analyze the following dataset...' })
})
console.log(await res.json())
```

<Note>All Apollo models currently return OpenAI-like shapes and are compatible with common SDKs. Choose a model based on your latency/cost tradeoffs.</Note>

## Daedalus-1 2B

* Model ID: `Daedalus-1-2B`
* Availability: Free (Beta)
* Context window: 128k tokens (code-optimized)
* Recommended use-cases: code generation, bug fixing, algorithmic reasoning, unit-test synthesis

### Example: Generate a Python function

```bash theme={null}
curl https://api.loom.aayanmishra.com/api/v1/responses \
  -H "Authorization: Bearer $LOOM_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"model":"Daedalus-1-2B","input":"Write a Python function that reverses a linked list and includes type hints and unit tests."}'
```

## Daedalus-1 8B

* Model ID: `Daedalus-1-8B`
* Availability: Free (Beta)
* Context window: 128k tokens (code-optimized)
* Recommended use-cases: large-scale code synthesis, complex debugging, instruction-following on codebases

### Example: Complex code transformation (Node)

```js theme={null}
const res = await fetch('https://api.loom.aayanmishra.com/api/v1/responses', {
  method: 'POST',
  headers: { 'Authorization': `Bearer ${process.env.LOOM_API_KEY}`, 'Content-Type': 'application/json' },
  body: JSON.stringify({ model: 'Daedalus-1-8B', input: 'Refactor this legacy function to use async/await and add error handling: <paste code>' })
})
console.log(await res.json())
```