Developer & Enterprise Knowledge
Frequently Asked Questions
Frequently Asked Questions
Everything You Need to Know
Essential technical, billing, and architectural guidance for integrating Softelv AI into production.
How do I integrate Softelv AI into my application?
Integration takes less than 60 seconds. Our endpoint is fully compatible with OpenAI SDKs.
Set
base_url="https://softelv.ai/v1" and pass your API key created in the developer console. No proprietary client libraries are required.
Which models are available and what is their inference speed?
We host high-throughput models optimized for low first-token latency:
• Llama 3.3 70B Fast: Ultra-fast flagship conversation model.
• DeepSeek R1 Distill: Deep mathematical and logical reasoning.
• Qwen 2.5 Coder 32B: High-performance software engineering.
• Llama 3.2 Vision: Multimodal visual question answering & OCR.
• FLUX.1 Schnell: Photorealistic high-resolution image synthesis in < 2 seconds.
• Llama 3.3 70B Fast: Ultra-fast flagship conversation model.
• DeepSeek R1 Distill: Deep mathematical and logical reasoning.
• Qwen 2.5 Coder 32B: High-performance software engineering.
• Llama 3.2 Vision: Multimodal visual question answering & OCR.
• FLUX.1 Schnell: Photorealistic high-resolution image synthesis in < 2 seconds.
Do you use customer API prompts to train models?
Strictly no. Softelv LLC upholds an absolute zero-training policy for all customer API traffic. Your prompts, embeddings, completions, and code are never stored for model training purposes.
Is real-time streaming supported?
Yes. All chat endpoints support standard Server-Sent Events (SSE). When you specify
stream: true, tokens are streamed incrementally with ultra-low latency.
What is the difference between Softelv AI and Nashmi AI?
Softelv AI (softelv.ai): The cloud developer API gateway allowing software teams to connect AI capabilities directly to their applications.
Nashmi AI (ai.softelv.com): The flagship consumer and business AI assistant chat application created by Softelv LLC.
Nashmi AI (ai.softelv.com): The flagship consumer and business AI assistant chat application created by Softelv LLC.
What corporate entity issues invoices?
All transactions and services are legally provided by SOFTELV LIMITED LIABILITY COMPANY, registered in Washington State (UBI 606 260 346) under Founder & Chairman Ahmad Montaser Mohammad Al-Amoudi.
How can I scale my rate limits or request dedicated capacity?
The free Starter plan includes 50 requests daily. Upgrading to Pro unlocks 120 RPM. For higher volume, dedicated GPU clusters, or custom SLAs, reach out directly to ai@softelv.com.