Introduction
Triangle LLM is the UK's premier AI language model, designed for enterprise applications with unmatched performance, data sovereignty, and GDPR compliance.
GDPR Compliant
Built with privacy by design, ensuring full compliance with UK and EU data protection regulations.
UK Data Centers
Your data never leaves the UK, with infrastructure hosted in secure British data centers.
Sub-50ms Latency
Lightning-fast responses with optimized edge computing across UK regions.
Quick Start
1. Get Your API Key
Sign up for a Triangle LLM account and generate your API key from the dashboard.
export TRIANGLE_API_KEY="your-api-key-here"2. Install the SDK
pip install triangle-llmfrom triangle_llm import TriangleLLM
client = TriangleLLM(api_key="your-api-key")
response = client.chat.completions.create(
model="triangle-pro",
messages=[
{"role": "user", "content": "Hello, Triangle LLM!"}
]
)
print(response.choices[0].message.content)Models
Triangle-Pro
Our most capable model for complex reasoning
Flagship
Capabilities
- • Advanced reasoning and analysis
- • Complex problem solving
- • Code generation and debugging
- • Research and writing
- • Mathematical computations
Specifications
- • 175B parameters
- • 32k context window
- • £3.50 per 1M tokens
- • ~50ms response time
- • 99.9% uptime SLA
Triangle-Fast
Optimized for speed and cost-effectiveness
Most Popular
Capabilities
- • Fast text generation
- • Simple Q&A and chat
- • Content summarization
- • Basic code assistance
- • Translation tasks
Specifications
- • 7B parameters
- • 8k context window
- • £0.80 per 1M tokens
- • ~25ms response time
- • 99.9% uptime SLA
Triangle-Vision
Multi-modal AI for text and image understanding
Multi-Modal
Capabilities
- • Image analysis and description
- • Visual question answering
- • Document understanding
- • Chart and graph interpretation
- • OCR and text extraction
Specifications
- • 90B parameters
- • 16k context window
- • £4.20 per 1M tokens
- • ~75ms response time
- • 99.9% uptime SLA
API Reference
Chat Completions
Create a chat completion with Triangle LLM models.
POST
/v1/chat/completionsRequest Body
{
"model": "triangle-pro",
"messages": [
{
"role": "system",
"content": "You are a helpful assistant."
},
{
"role": "user",
"content": "Hello!"
}
],
"temperature": 0.7,
"max_tokens": 1000,
"stream": false
}Response
{
"id": "chatcmpl-123",
"object": "chat.completion",
"created": 1677652288,
"model": "triangle-pro",
"choices": [
{
"index": 0,
"message": {
"role": "assistant",
"content": "Hello! How can I help you today?"
},
"finish_reason": "stop"
}
],
"usage": {
"prompt_tokens": 9,
"completion_tokens": 12,
"total_tokens": 21
}
}Streaming
Stream responses in real-time for better user experience.
const response = await client.chat.completions.create({
model: 'triangle-pro',
messages: [
{ role: 'user', content: 'Write a story about AI' }
],
stream: true
});
for await (const chunk of response) {
const content = chunk.choices[0]?.delta?.content || '';
process.stdout.write(content);
}Best Practices
Prompt Engineering
- • Be specific and clear in your instructions
- • Use system messages to set context and behavior
- • Provide examples for complex tasks
- • Break down complex requests into steps
- • Use temperature settings to control creativity
Performance Optimization
- • Use Triangle-Fast for simple tasks to reduce costs
- • Implement caching for repeated requests
- • Use streaming for long responses
- • Batch multiple requests when possible
- • Monitor token usage and optimize prompts
Security & Privacy
- • Never include API keys in client-side code
- • Use environment variables for sensitive data
- • Implement rate limiting in your applications
- • Validate and sanitize user inputs
- • Monitor API usage for unusual patterns
Support & Community
Get Help
Multiple ways to get support when you need it
Resources
Additional resources and tools