Skip to main content
DeepMask is built around the principle that no single model is best for every task. You can switch models at any point — including mid-conversation — without losing your context. Use the tabs below to find a good starting point for what you’re working on.
Not sure? Choose DeepMask D1. It scores each message and passes it to a suitable model for you.
These models suit code generation, debugging, refactoring, and longer development tasks that use tools.
You can switch models at any point in a conversation. If a response isn’t working well, select a different model from the model picker and continue — DeepMask carries your context forward automatically.

Model capability quick reference

A model with built-in reasoning (sometimes called “thinking”) works through a problem step by step before writing its final answer. This takes longer but gives better results for math, multi-step logic, planning, and any task where intermediate steps matter.Models with built-in reasoning include the Claude 5.5 and 5 families, Opus 4.5–4.8, the GPT-6 series, GPT-o3 Mini, most Gemini 3 Flash models, Qwen3.8, Qwen3.8 27B (evroc), Qwen3-VL 30B, Qwen3 (StackIT), Kimi K3, Kimi K2.6 (evroc), Kimi K2 (DeepMask), GLM-5.3, GLM 5.3 (evroc), GLM-5, DeepSeek V4 Flash, DeepSeek V3.2, MiniMax M3, MiniMax M2.5, and Nemotron 3 Super. The overview table shows each model.
These models can read images such as screenshots, photos, charts, and scanned pages:
  • Anthropic: all Opus, Sonnet, and Haiku models
  • OpenAI: GPT-6.1 Sol, GPT-6 Sol, GPT-6 Luna, GPT-5.4, GPT-5.2, GPT-4.1, GPT-4o
  • Google: all Gemini 3 Flash models, Gemma 4 31B, Gemma 3 27B (StackIT)
  • Alibaba: Qwen3.8 27B (evroc), Qwen3.5 397B, Qwen3.5 122B, Qwen3-VL 30B, Qwen3 (StackIT)
  • Moonshot AI: Kimi K3, Kimi K2.6 (evroc), Kimi K2.5
  • MiniMax: MiniMax M3
  • Mistral AI: Mistral Large 3
All other models — including GPT-o3 Mini, GPT-OSS 120B, the DeepSeek and GLM models, Qwen3.8, Qwen3 235B (2507), Kimi K2 (DeepMask), Kimi K2.7 Code, MiniMax M2.5 and M2.1, Nemotron 3 Super, and DeepMask D1 and D2 — work with text and documents only.
Tool use (also called function calling) lets a model call connectors, web search, and skills during a conversation. DeepMask exposes tools through its MCP connector framework, which supports Microsoft Enterprise, OneDrive, Outlook, SharePoint, Salesforce, and more.Every model in DeepMask supports tool use except Gemma 3 27B (StackIT).
The context window is the maximum amount of text (measured in tokens, where 1 token ≈ 0.75 words) that a model can read in a single conversation. Larger context windows let you work with longer documents, more conversation history, and bigger codebases.The overview table lists the exact limit for each model. Use DeepMask Projects to keep files and instructions available across sessions, whichever model you choose.