Skip to main content
Gemini 3.7 Flash (Google DeepMind) is the new Flash workhorse in Tess — built for coding, production agents, and knowledge work, with intro pricing at half of Gemini 3.6 Flash through 31 December 2026.

​
What changed vs Gemini 3.6 Flash

Native reasoning, tools (function calling / MCP, search, computer use), and vision. Strong fit for multi-agent coding subagents in production.

​
Pricing (Tess credits)

Values follow Models and Costs (credits per 100 tokens). Intro rate through 31 December 2026:
Screenshot placeholder — model picker: Capture the chat model selector with Gemini 3.7 Flash selected.

​
Ideal use cases in Tess

  1. Production coding agents (long-horizon tasks, +16 pts DeepSWE vs 3.6)
  2. Multi-step agent workflows — planning, tool loops, error recovery
  3. Knowledge work — financial documents and business reports
  4. Web development against visual references
  5. Low-cost subagents during the intro pricing window
  6. Multimodal turns (text + image + audio + video)
Best practices
  • Prefer 3.7 Flash over 3.6 Flash for coding agents and complex multimodal jobs — same context, lower cost, higher coding quality.
  • Keep 3.5 Flash-Lite for high-volume extraction, classification, and summarization.
  • Intro pricing is time-boxed: plan volume before 31 Dec 2026 if cost is the main reason to switch.
See also: Gemini Flash · Models and Costs · Gemini 3.7 Flash (Google).