Google’s Gemini 3.5 Flash scores within two points of Anthropic’s flagship at roughly one third of the price

Google’s Gemini 3 triggered a Code Red at OpenAI last December and briefly seized the frontier narrative. Roughly half a year later, the AI model race had mostly become a two-company contest between Anthropic and OpenAI with Claude and GPT trading the top spots on independent leaderboards. Meanwhile, Google’s developer tools steadily lost ground to Claude Code and OpenAI’s Codex. Gemini 3.5 Flash, released today at Google I/O, complicates that story. On the independent Artificial Analysis Intelligence Index, it scores 55, within two points of Claude Opus 4.7 and five points of GPT-5.5, at roughly a third of the cost per token.

Google has also made 3.5 Flash available in AI Mode in Google Search

Google says 3.5 Pro is currently in testing and will be available next month. If Flash already scores 55 on the composite intelligence index, Pro could land in the frontier range, putting it in direct competition with Claude Opus 4.7 and GPT-5.5. Google also may undercut rivals on cost. Flash is priced at $1.50 input and $9.00 output per million tokens, while GPT-5.5 launched at $5.00 input and $30.00 output, a 2x increase over GPT-5.4. If Pro follows Google’s pricing pattern, it could deliver frontier-class intelligence for less than what Anthropic and OpenAI charge. When Gemini 2.5 Pro was released, it undercut essentially all mid-2025 frontier rivals on cost.

Whether Gemini 3.5 Pro can gain the LLM crown is a different matter. Google’s models have tended to edge out rivals in some areas (multimodal understanding, speed, context length) while lagging in others (coding workflows, agentic reliability, developer tooling). Gemini 3.1 Pro launched in February 2026 to benchmark scores that looked strong on paper. Community feedback, however, was more qualified, with users reporting uneven behavior inside third-party harnesses and a more clinical conversational tone. Google positioned 3.5 Flash as combining “frontier intelligence with action” and called it “a major leap forward in building more capable, intelligent agents.”

The Flash release arrives at a moment when the two companies commanding premium pricing are both dealing with user pushback. Anthropic released Claude Opus 4.7 on April 16, and within 48 hours developer forums and independent analyses documented a split in reception. The core complaints centered on token burn roughly 1.5 to 3x higher than Opus 4.6 in practice, along with confidently broken output on long runs. Developers who reported Opus 4.7 issues were simultaneously describing successful workflows with OpenAI’s Codex or returning to Opus 4.6 as a pinned version. OpenAI, meanwhile, doubled its flagship API pricing with GPT-5.5.