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

# fastgpt

> Complete reference for *kagi* fastgpt command - get quick answers using Kagi's FastGPT API.

# `kagi fastgpt`

Answer queries using Kagi's FastGPT API for quick, factual responses.

## Synopsis

```bash theme={null}
kagi fastgpt [OPTIONS] <QUERY>
```

## Description

The `kagi fastgpt` command queries Kagi's FastGPT API, which provides quick, factual answers optimized for speed and accuracy. Unlike the Assistant which maintains conversation threads, FastGPT is designed for single-question answering.

**Key characteristics:**

* Fast response times
* Factual, concise answers
* No conversation state
* API-based (consumes credit)

## Authentication

**Required:** `KAGI_API_TOKEN`

FastGPT currently uses Kagi's legacy `/api/v0` API with `Bot` auth. It requires API access and consumes API credit per request.

## Arguments

### `<QUERY>` (Required)

The question or query to answer.

**Example:**

```bash theme={null}
kagi fastgpt "What is the capital of France?"
kagi fastgpt "Explain machine learning in simple terms"
```

## Options

### `--cache <BOOL>`

Enable or disable response caching.

**Values:** `true`, `false`
**Default:** `true`

When enabled, identical queries may return cached responses for faster results and reduced API usage.

**Example:**

```bash theme={null}
kagi fastgpt "What is Rust?" --cache false
```

### `--web-search <BOOL>`

Enable web search augmentation.

**Values:** `true`
**Default:** `true` (web search enabled)

**Note:** Kagi's documentation notes that values other than `true` for `--web-search` are currently unsupported in practice.

### `--local-cache`

Store and reuse the response in the local cache under `~/.cache/kagi-cli` or `KAGI_CACHE_DIR`.

### `--cache-ttl <SECONDS>`

Override the local cache TTL. FastGPT defaults to 3600 seconds when `--local-cache` is enabled.

## Output Format

```json theme={null}
{
  "meta": {
    "id": "req-123",
    "node": "us-east",
    "ms": 275
  },
  "data": {
    "output": "The answer to your question...",
    "tokens": 57,
    "references": []
  }
}
```

### Fields

| Field             | Type    | Description                          |
| ----------------- | ------- | ------------------------------------ |
| `meta`            | object  | Request metadata from the API        |
| `data.output`     | string  | FastGPT's answer                     |
| `data.tokens`     | integer | Token count reported by the API      |
| `data.references` | array   | Supporting references when available |

## Examples

### Basic Queries

```bash theme={null}
# Factual question
kagi fastgpt "What is the speed of light?"

# Explanation
kagi fastgpt "How does photosynthesis work?"

# Comparison
kagi fastgpt "What's the difference between TCP and UDP?"
```

### Processing Output

```bash theme={null}
# Extract just the answer
kagi fastgpt "What is Docker?" | jq -r '.data.output'

# Save response
kagi fastgpt "Explain blockchain" > blockchain-explanation.json

# Use in scripts
ANSWER=$(kagi fastgpt "What is $TECHNOLOGY?" | jq -r '.data.output')
echo "Technology: $TECHNOLOGY"
echo "Explanation: $ANSWER"
```

### Quick Reference Tool

```bash theme={null}
#!/bin/bash
# quick-answer.sh

TERM="$1"
[ -z "$TERM" ] && { echo "Usage: quick-answer.sh <term>"; exit 1; }

echo "🔍 Quick answer for: $TERM"
echo ""
kagi fastgpt "What is $TERM?" | jq -r '.data.output'
```

## Exit Codes

| Code | Meaning                   |
| ---- | ------------------------- |
| 0    | Success - answer received |
| 1    | Error - see stderr        |

**Common errors:**

* Missing API token
* Insufficient API credit
* Rate limiting
* Network error

## Comparison with Assistant

| Feature           | FastGPT      | Assistant                  |
| ----------------- | ------------ | -------------------------- |
| **Token**         | API          | Session                    |
| **Speed**         | Faster       | Slightly slower            |
| **Conversations** | Single query | Threaded                   |
| **Cost**          | API credit   | Included with subscription |
| **Use case**      | Quick facts  | Complex discussions        |

## When to Use FastGPT

**Good for:**

* Quick factual questions
* Simple explanations
* Single-turn queries
* Automated workflows
* Cost-sensitive operations

**Not ideal for:**

* Multi-turn conversations
* Complex reasoning chains
* Context-dependent queries
* Creative writing

## Best Practices

### Clear Questions

```bash theme={null}
# Good - specific
kagi fastgpt "What are the main differences between Python 2 and Python 3?"

# Less effective - vague
kagi fastgpt "Python"
```

### Error Handling

```bash theme={null}
if ANSWER=$(kagi fastgpt "Question" 2>/dev/null | jq -r '.data.output'); then
  echo "Answer: $ANSWER"
else
  echo "Failed to get answer - check API credit"
fi
```

### Caching Strategy

```bash theme={null}
# For frequently asked questions, caching reduces cost
kagi fastgpt "What is REST?" --cache true

# For time-sensitive queries, disable caching
kagi fastgpt "What is the current weather in Tokyo?" --cache false
```

## See Also

* [assistant](/commands/assistant) - Conversational AI
* [summarize](/commands/summarize) - Content summarization
* [Authentication](/guides/authentication) - API token setup

***

*Last updated: March 2026*