DeepSeek R2

At a glance

What DeepSeek R2 does

Tool use, Streaming, Structured output, Extended reasoning, Chain-of-thought, Function calling, Codeforces 2318

Best for

Hard reasoning at 1/20th o3 cost, math (AIME 83.2%, MATH 98.1%), competitive programming, research-grade analysis on a budget

Works well with

Conflicts & caveats

• Google Antigravity is built around Gemini 3 Pro; Claude and recent GPT-5 models also work, but DeepSeek R2 is not in the supported model list.
• Kiro's spec mode is tuned for Claude models — DeepSeek R2 may work but spec-driven flows are best with Claude.
• OpenAI Codex (VS Code) is tied to OpenAI's GPT-5.1 Codex family — DeepSeek R2 is not selectable inside this integration.
• Gemini CLI is locked to Google Gemini models — your LLM is DeepSeek R2.
• Claude Code only works with Claude models — your LLM is DeepSeek R2.
• Devin uses Cognition's proprietary model internally — your selected DeepSeek R2 will be ignored.
• OpenAI Codex (Cloud) runs GPT-5.1 Codex internally — your selected DeepSeek R2 will be ignored.
• Jules runs Gemini models internally — your selected DeepSeek R2 will be ignored.
• The Claude Code SDK is tuned for Claude models — DeepSeek R2 may work via the model adapter but loses Claude-specific features (thinking, prompt caching, sub-agents).
• Lovable uses a fixed internal model — your selected DeepSeek R2 will be ignored.
• Replit Agent 3 uses its own model selection internally — your selected DeepSeek R2 will be ignored.