Three new AI models are now available in Pickaxe!
Gemini 3.7 Flash
• Fast, efficient performance for everyday AI tools and high-volume workflows
• Stronger coding, instruction-following, and multi-step reasoning
• Multimodal support for working with text, images, audio, video, and large documents
• A great all-around choice when you want a balance of speed, intelligence, and cost
Google positions Gemini 3.7 Flash as the high-efficiency “workhorse” of the Gemini 3 family, with customizable reasoning and a 1M-token context window.
Grok 4.6
• Built to stay on track during longer, more complex tasks
• Strong for research, coding, analysis, and multi-step agent workflows
• Improved first drafts for interactive apps and visually polished projects
• Better self-testing and verification as it works through a task
xAI designed Grok 4.6 specifically for long-running agents, ambitious coding projects, knowledge work, and stronger interactive and visual outputs
DeepSeek V4 Pro 0813
• Deep reasoning for difficult or highly technical prompts
• Enhanced coding, tool use, and production-grade agent performance
• Adjustable reasoning effort for balancing speed and depth
• A strong option for complex automation and agent-driven workflows
DeepSeek’s latest V4 Pro release focuses heavily on stronger real-world agent capabilities and supports low, high, and max reasoning effort levels.
Try them in your existing Pickaxes, compare their results, and let us know which one becomes your new favorite 