Custom n8n node for integrating with the AI/ML API platform (AIMLAPI) to interact with LLMs and multimodal AI models such as chat completion endpoints.
npm install n8n-nodes-aimlapin8n-nodes-aimlapi
bash
npm install n8n-nodes-aimlapi
`
Then restart your n8n instance.
---
Configuration
$3
To use this node, you must configure AI/ML API credentials:
1. Sign up at https://aimlapi.com/
2. Go to your dashboard and generate an API key
3. In n8n, create new credentials of type AI/ML API
4. Paste your API key
5. Save and use in the node
📚 You can also refer to the provider documentation for detailed API specs.
$3
- Operation – Select the modality (Chat Completion, Image Generation, Audio Generation, Video Generation, Speech Synthesis, Speech Transcription, or Embedding Generation).
- Model – Pick a model from the filtered list (dependent on operation) or provide an ID via expression.
#### Chat Completion
- Use Message List – Switch between a single prompt and a structured list of chat messages.
- Prompt/Messages – Provide either one prompt or multiple messages. Each structured message selects a Role (assistant, system, or user) from the dropdown, or choose Custom… to enter another role such as tool, plus the text Content. When using the tool role, fill in the required Tool Call ID with the identifier returned alongside the assistant tool invocation so the API can pair the response correctly.
- Extract From Response – Decide what to return (Text Only, Assistant Messages, Choices Array, or Full Raw JSON).
- Options – Fine-tune completions with Temperature, Top P, Max Tokens, Presence Penalty, Frequency Penalty, and Response Format.
#### Image Generation
- Prompt – Describe the image you want the model to create.
- Extract From Response – Return the first/all URLs, first/all base64 images, or the raw payload.
- Image Options – Configure Background, Image Count, Negative Prompt, Quality, Response Format, Size, and Style.
#### Audio Generation
- Prompt – Describe the audio clip you want to generate.
- Extract From Response – Return audio URLs, base64 blobs, or the raw payload.
- Audio Options – Adjust Audio Format, Duration, Mode, Negative Prompt, Prompt Strength, Seed, Reference Audio URL, Instrument, and CFG Scale.
#### Video Generation
- Prompt – Describe the scene for the generated video.
- Extract From Response – Return the first/all video URLs or the raw payload.
- Video Options – Control Aspect Ratio/Ratio, Duration, Kling Types & Image Lists, Google Veo Tail Images & Prompt Enhancement, Alibaba Resolution/Watermark/Prompt Expansion, MiniMax Prompt Optimizer & First Frame, Runway Last Image, plus legacy toggles like Negative Prompt, Prompt Strength, Reference Media, Background Audio, CFG Scale, and Seed.
#### Speech Synthesis
- Text – Content that should be spoken aloud.
- Extract From Response – Return an audio URL, base64 payload, or the raw response.
- Speech Options – Configure Voice, Audio Format, Sample Rate, Speaking Rate, and Style.
#### Speech Transcription
- Binary Property – Name of the binary property that contains the audio file.
- Extract From Response – Return transcript text, segment breakdowns, or the raw payload.
- Transcription Options – Provide Language, Prompt, Response Format, and Temperature.
#### Embedding Generation
- Input Text – Text that should be converted into an embedding vector.
- Extract From Response – Return the first vector, all vectors, or the raw payload.
- Embedding Options – Configure Encoding Format and User Identifier metadata.
---
Usage Example
1. Add a Manual Trigger
2. Add the AI/ML API node
3. Configure credentials
4. Set a model (e.g. gpt-3.5-turbo)
5. Enter prompt: What is the capital of France?
6. Select Extract: Text Only
7. Execute — you’ll get Paris
---
Troubleshooting
- If the node doesn't show up, try restarting your instance
- If the icon or name doesn't render, make sure you ran npm run build && gulp build:icons
- To debug model list: ensure credentials and /models` endpoint return valid JSON