Flexible. Cost-Effective. Controlled.

Any AI model.
Cloud or local. Your choice.

Not locked to one AI vendor. Connect OpenAI, Anthropic, Ollama, or your favorite LLM. Setup takes 3 minutes. You pay directly to your LLM provider – transparent and traceable.

Start integration Watch guide

Supported AI Providers

All major cloud providers + local open-source models

OpenAI

Modelle: GPT-4o, GPT-4o mini
Speed: Very fast
Kosten: ~$0.01/report
Setup: 2 minutes
Datenschutz: Cloud (USA)

Reliable, fast, best-in-class quality

Anthropic

Modelle: Claude 3.5 Sonnet, Claude 3 Haiku
Speed: Fast
Kosten: ~$0.02/report
Setup: 2 minutes
Datenschutz: Cloud (USA)

Precise, security-focused, longer context

Google

Modelle: Gemini Pro
Speed: Fast
Kosten: Variable
Setup: 2 minutes
Datenschutz: Cloud (Google)

Multimodal, good for vision tasks

Ollama (local)

Modelle: Llama 2, Mistral, Neural Chat
Speed: Medium
Kosten: Free
Setup: 10 minutes
Datenschutz: 100% local

Completely on your hardware, no API access

LM Studio (local)

Modelle: Any Hugging Face model
Speed: Variable
Kosten: Free
Setup: 15 minutes
Datenschutz: 100% local

GUI for local models, easy to use

Custom OpenAI-API

Modelle: Any OpenAI-compatible endpoint
Speed: Variable
Kosten: Your costs
Setup: 5 minutes
Datenschutz: Depends on provider

For custom LLMs or on-premise solutions

Why BYO LLM?
Bring Your Own Model

Full Cost Control

You pay your LLM provider directly. No hidden fees, no markup fees, no Bygmind subscriptions.

No Vendor Lock-in

Today OpenAI, tomorrow Anthropic, day after tomorrow Ollama locally. Switch without data loss or re-setup.

Highest Data Security

All data can be processed locally. Your documents never leave your infrastructure. GDPR and data protection by design.

Optimal Model Choice

Pick the best model for your use case – not what Bygmind recommends. Speed, cost, quality: you decide.

Transparent Billing

See exactly what your reports cost. OpenAI pricing is public. Ollama costs $0.

Cost Examples per Report

With typical project (5 photos, 2 min audio)

OpenAI (GPT-4o-mini)
~$0.01
Verarbeitung: 30 seconds
Best für: Budget-conscious, good speed/cost balance
Anthropic (Claude Haiku)
~$0.015
Verarbeitung: 45 seconds
Best für: Precision required, higher quality
Google Gemini
~$0.001
Verarbeitung: 60 seconds
Best für: Very cost-effective, new to market
Ollama (local on M1 Mac)
$0
Verarbeitung: 3–5 minutes
Best für: Free, maximum data security, slower than cloud

3-Minute Connection Guide

1

Get API Key

OpenAI, Anthropic, Google – each provider has a dashboard with API keys. Sign up, create a key.

Go to OpenAI Dashboard →
2

Enter Key in Bygmind

Open Bygmind Settings → AI Agent → Select Provider → Paste API Key. Save.

See screenshot →
3

Create Test Report

Upload documentation, click "Create Report", done. Everything works. You pay only for actual usage.

Watch video guide →

Local AI Models Setup

Ollama (macOS/Linux)

brew install ollama && ollama run mistral. Then enter endpoint (localhost:11434) in Bygmind. Done.

LM Studio (macOS/Windows)

GUI-based. Download model from Hugging Face, start it, connect Bygmind. No terminal knowledge required.

Requirements for Local Models

Min. 8GB RAM recommended. Llama 2 (~13B) needs ~10GB. Mistral ~7B ~6GB. On Apple Silicon: ~110x real-time speed (like offline transcription).

🔗

Connect Any AI Model

OpenAI, Anthropic, Google, Ollama – no lock-in, unlimited flexibility.

💰

Transparent Costs

Pay the provider directly. ~$0.01 per report with OpenAI or free with Ollama.

⚡

Quick Integration

3-minute setup. Paste API key, done. No complex configuration.

🛡️

Local or Cloud – Your Choice

Maximum data security with local LLM or proven cloud quality with OpenAI.

🔀

Switch Anytime

No vendor lock-in. Today OpenAI, tomorrow Ollama – all your data stays.

⚙️

Custom Parameters

Adjust temperature, top-K and other parameters per model – full control.

No. You can configure multiple API keys in Bygmind and then select per project. One project can run on OpenAI, the next on Anthropic – without needing multiple apps or logins.
Your API keys are encrypted locally on your device. They are not transmitted to Bygmind servers. Bygmind only sends your documentation to the AI provider of your choice – not to Bygmind.
OpenAI: fast, reliable, cloud-based (~$0.01/report). Anthropic: more precise, context-sensitive, cloud-based (~$0.02/report). Ollama local: free, 100% private, but slower (3–5 min). Choose based on your needs: speed, cost, or data security.
Yes. You can run multiple Ollama instances on different servers or ports. Bygmind then connects to the endpoint of your choice. Ideal for load balancing or failover.
All already-created reports remain unchanged. New reports are created with the new model. No data migration required – just switch the provider in settings.
Yes, with most providers. OpenAI and Anthropic offer spending limits and alerts. You can also create API keys with separate limits per project. Full cost control.

Choose your
perfect AI model.

Three minutes setup, full control. Get started free.

Connect Now

Get in Touch

Have questions about Bygmind? We're here to help.

Your data is secure. GDPR compliant. We'll respond as soon as possible.