---
title: "通过 RunAPI 在爱马仕 (Hermes Agent) 中使用 Ideogram V3 — 图像 API 指南"
url: "https://runapi.ai/zh-CN/hermes-ideogram-v3.md"
canonical: "https://runapi.ai/zh-CN/hermes-ideogram-v3"
locale: "zh-CN"
model: "ideogram-v3"
---

# 在 Hermes Agent 中使用 Ideogram V3。

Ideogram V3 生成的图像具有行业领先的文本渲染能力 — 在海报、logo 和标识上呈现其他图像模型难以胜任的可读排版。它提供三个端点（text_to_image、edit_image、remix_image）和一个 turbo 速度层级，能以更低成本将生成时间缩短一半。Hermes Agent 通过用于聊天的同一个 RunAPI 自定义 provider 和 API 密钥调用它，无需安装额外插件。

## API example

```bash
curl -X POST https://runapi.ai/v1/text_to_image \
  -H "Authorization: Bearer $RUNAPI_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "ideogram-v3-text-to-image",
    "prompt": "A minimalist cafe menu poster, cream background, the text FRESH BREWED written in elegant serif font at the top, three coffee cup illustrations below, warm earth tones",
    "style": "design",
    "aspect_ratio": "3:4",
    "rendering_speed": "balanced"
  }'

```

### Response

```json
{
  "task_id": "tsk_abc123",
  "status": "pending",
  "model": "ideogram-v3-text-to-image"
}

```

## How it works

1. **Configure RunAPI** — Set the RUNAPI_API_KEY environment variable in your shell profile. If the custom:runapi provider is already configured in Hermes Agent for chat, the same key works for Ideogram V3 — no additional setup needed.
2. **Call Ideogram V3** — Send a POST request to the text_to_image endpoint with model set to ideogram-v3-text-to-image. Include your prompt with any text you want rendered in the image. Set style to design for poster and logo work, or realistic for photographic output. Use rendering_speed turbo for faster results at lower cost.
3. **Get the result** — The API returns a task_id immediately. Poll the task status endpoint until the status changes to completed, then retrieve the output image URL from the response. Request up to 4 images per call with output_count.

## Parameters

| Parameter | Type | Description |
|-----------|------|-------------|
| `model` | `string` | Required. ideogram-v3-text-to-image, ideogram-v3-edit, or ideogram-v3-remix. |
| `prompt` | `string` | Text description of the desired image. Include specific text you want rendered. |
| `style` | `string` | Optional. auto, general, realistic, or design. Use design for posters and logos. |
| `rendering_speed` | `string` | Optional. turbo (fastest, lowest cost), balanced (default), or quality (highest detail). |
| `aspect_ratio` | `string` | Optional. 1:1, 3:4, 9:16, 4:3, or 16:9. |
| `output_count` | `integer` | Optional. Number of images to generate per request. 1 to 4. |
| `source_image_url` | `string` | Required for edit and remix endpoints. URL of the source image. |

## FAQ

### How does Ideogram V3 handle text in images compared to other models?

Ideogram V3 is specifically optimized for typography rendering. It can produce readable multi-line text, stylized headings, and logo-quality lettering that models like Flux or Imagen often garble. If your use case requires accurate text on posters, signs, or product labels, Ideogram V3 is the strongest option in the RunAPI catalog.

### What is the difference between turbo, balanced, and quality rendering speeds?

Turbo generates the fastest at roughly half the cost of balanced. Quality produces the highest-detail output but takes longer. Balanced is the default middle ground. All three tiers use the same underlying Ideogram V3 model — the difference is compute time and fidelity, not model version.

### How do I configure the custom:runapi provider in Hermes Agent for Ideogram V3?

Add a custom provider with base_url https://runapi.ai/v1 and your RUNAPI_API_KEY in the Hermes Agent provider config. Once configured, call any RunAPI endpoint — text_to_image, edit_image, remix_image — through the same provider. Run hermes doctor to verify the connection.

### Can I use remix mode to transfer the style of one image to another?

Yes. Use the remix_image endpoint with model set to ideogram-v3-remix. Pass source_image_url with the style reference image, then describe the new composition in the prompt. The output inherits the visual style of the source while following the text prompt for content. You can also set style to design, realistic, or general to further guide the output.

### How does Hermes Agent handle Ideogram V3's text rendering parameters?

Hermes Agent passes your prompt (including text content and placement instructions) to the RunAPI Ideogram V3 endpoint. The model handles text rendering natively. For best results, clearly specify the text to render and its position in the prompt.

### Is Ideogram V3 worth the cost compared to free alternatives?

For text-heavy designs -- logos, posters, signage -- Ideogram V3 saves significant time over alternatives that require manual text correction. The turbo tier cuts cost in half versus balanced mode while using the same model. For images without text, other models like Flux 2 or Imagen 4 may offer better value.


## Links

- [Hermes Agent 配置指南 →](https://runapi.ai/zh-CN/hermes-agent)
- [Ideogram V3 模型 →](https://runapi.ai/zh-CN/models/ideogram-v3)
- [Model catalog](https://runapi.ai/zh-CN/models)
- [API docs](https://runapi.ai/zh-CN/docs)