Help Center

The system should prompt you for this if you have not already configured it.

If you need to do this manually on the desktop app:

• Go to Settings → LLM Backend.
• Either paste your OpenRouter API key into the API key field, or start Ollama to use a local backend.

To run Ollama as a server endpoint, allowing other applications or clients to interact with your local LLMs, follow these steps:

• Install Ollama: If you haven't already, download and install Ollama on your system. This will set up the necessary components and, by default, start the Ollama server in the background, typically listening on port 11434.
• Verify server status (optional): Open a terminal or command prompt and run:

  ollama list

  This command lists the models you have downloaded and also indicates if the Ollama server is active.

• Manually start (if needed): If the server is not running, or if you need to restart it, start a new instance:

  ollama serve

  Ensure no other Ollama server instances are running to avoid port conflicts.

• Access the API: Once the Ollama server is running, you can interact with it via its REST API at http://localhost:11434 (or the IP address of the machine running Ollama if accessing remotely).
• Interact with models (examples):
  • Generate completion:

    curl http://localhost:11434/api/generate \
      -H "Content-Type: application/json" \
      -d '{ "model": "llama3", "prompt": "Tell me a joke.", "stream": false }'

  • Chat with a model:

    curl http://localhost:11434/api/chat \
      -H "Content-Type: application/json" \
      -d '{ "model": "llama3", "messages": [{"role": "user", "content": "Hi!"}] }'

  • List available models:

    curl http://localhost:11434/api/tags

• Important considerations:
  • Security: If exposing your Ollama server beyond your local machine, implement security measures like a reverse proxy with HTTPS and authentication, as Ollama's API itself lacks built-in authentication.
  • Resource management: Monitor CPU/GPU and memory usage, especially when serving models remotely or under heavy load.
  • Streaming: For real-time applications like chat interfaces, utilize the "stream": true option in API requests to receive tokens as they are generated.

Click Start Now on the homepage and follow the prompts.

Log in to your dashboard and open the billing tab.

In your dashboard, select API Keys and click regenerate. Note that you will need to update your API key in your desktop app manually for security reasons.

No, there are no usage limits on how much you can use G6Solver's free or paid plans. However, cost of paid LLM backend providers is separate to the cost G6solver.

No, we do not provide support for any other backend providers at this stage; however, we may consider doing so if there is enough interest in future.

Visit the News page for the latest releases, system updates and important announcements.

G6 Solver is like giving your AI assistant a brain upgrade. While regular AI chatbots (like ChatGPT) forget everything between conversations, G6 Solver creates AI agents that remember what you've worked on, learn from mistakes, and get better over time.

Think of it this way: regular AI is like talking to someone with short-term memory loss who's really smart but forgets everything you just discussed. G6 Solver is like having a brilliant research assistant who takes detailed notes, remembers your preferences, and builds on previous conversations to solve complex problems step-by-step.

G6 Solver fixes the biggest frustrations people have with AI assistants:

• Forgetfulness: Regular AI forgets your previous conversations. G6 Solver remembers everything—like having a research partner who keeps detailed project notes.
• Inconsistent quality: Sometimes AI gives brilliant answers, sometimes nonsense. G6 Solver includes built-in quality checks, like having an editor review every response.
• Can't learn: Most AI makes the same mistakes repeatedly. G6 Solver learns from errors—if it misunderstands your writing style once, it won't make that mistake again.
• Poor at complex tasks: Regular AI struggles with multi-step projects like "research this topic, then create a presentation, then prepare talking points." G6 Solver can handle complex, multi-day projects autonomously.
• Safety concerns: G6 Solver includes built-in safeguards to prevent harmful or inappropriate responses, like having a compliance officer review everything.

G6 Solver is perfect for:

• Software developers who need AI that can remember coding patterns across projects and learn their preferred coding style.
• Researchers and analysts working on complex, multi-week projects requiring consistent context and quality.
• Business teams who need reliable AI assistance with governance controls (ensuring responses meet company standards).
• Privacy-conscious users who want powerful AI without sending sensitive data to external servers.
• Organizations that need flexibility to use different AI models (not locked into one vendor like OpenAI or Anthropic).

G6 Solver works as a smart layer on top of any AI model (like ChatGPT, Claude, or others). Imagine it as having two assistants working together:

1. Chat Assistant: The friendly interface you talk to (like the chat window you're used to).
2. Work Assistant: The behind-the-scenes agent that actually does the heavy lifting, remembers everything, and ensures quality.

Instead of being limited to one AI company's model, G6 Solver can use any AI model through services like OpenRouter (which gives access to dozens of different AI models) or even AI models running locally on your computer. It's like having a universal remote control for AI models.

G6 Solver is vendor-agnostic and works with any LLM through OpenRouter or local Ollama deployments, giving you flexibility in model selection.

• Operating Systems: Windows, macOS, Linux, or Web Browser.
• Processor: 64-bit processor required.
• Memory: 4GB RAM minimum (8GB recommended).
• Storage: Approximately 2GB available space.
• Desktop version: Single-device license enforcement.

• Free Demo: Browser-based with limited functionality.
• Desktop Subscription: $20 AUD/month with full features.

• Demo: Browser-based, no data persistence, limited memory, reduced functionality, no login required.
• Desktop: Full features, persistent memory, project management, advanced configuration, secure local deployment.

Yes! You can try G6 Solver immediately through our free browser-based demo with no signup required.

1. Try the free demo directly in your browser to explore basic functionality.
2. Subscribe for the desktop application to access full features.
3. Receive your unique secure download link and single-device license.
4. Follow the setup guide for your operating system.

We provide comprehensive support through:

• Help Center and Documentation.
• API Reference guides.
• Community forums.
• System Status monitoring.
• Email support for general, technical, and business inquiries.

• Desktop: All data is stored locally on your device with encryption at rest. Data is not sent to G6 unless you explicitly share it for support purposes.
• Demo: No data persistence beyond your browser session.
• Security measures: TLS encryption in transit, AWS hosting with Cloudflare protection, secure coding practices, regular audits, and rate limiting.

• Desktop licenses are restricted to single-device use.
• Prohibited activities include reverse engineering, sharing licenses, circumvention attempts, and unlawful use.
• Demo has functional limitations but no usage restrictions.

G6 Solver offers unique capabilities including:

• Persistent, hierarchical memory architecture.
• Explicit learning from experience.
• Hybrid reasoning and program synthesis.
• Built-in quality metrics (accuracy, trustworthiness, precision, confidence).
• Governance and guardrail systems.
• Mathematical proof and precision features.
• Modular domain adaptation.
• Complete vendor independence.

Yes, G6 Solver is specifically designed for advanced multistep planning and autonomous problem-solving. It maintains context across complex workflows and learns from each interaction to improve future performance.

Contact our support team through:

• Technical support email for desktop application issues.
• Help Center for self-service troubleshooting.
• Community forums for user discussions.
• System Status page for service updates.

We have a responsible disclosure policy in place. Please contact our security team directly for any security-related concerns at [email protected].