Google DeepMind publishes technical paper on Gemini 3.7 Flash Thinking. Benchmark results show the model reaches 68.0% on SWE-bench Verified across a native 1M context window at 1/20th the token cost of traditional frontier models.

Key Takeaways

  • ✓Google DeepMind publishes official technical whitepaper for Gemini 3.7 Flash Thinking;
  • ✓1M native context with adaptive thinking delivers robust multi-file code refactoring;
  • ✓Scores 68.0% on SWE-bench with disruptive pricing of $0.15 / $0.60 per 1M tokens.