Physix Frontier · News Briefing Card (TMTPost · Nov 20, 2026)

Google Ships Fourth Gemini 4 Model in 49 Days, Opens Genie 4

KEY FACTS

  • Google released Gemini 4 Neon on November 18, the fourth Gemini 4 model in 49 days.
  • Neon scored 61 on the Artificial Analysis composite benchmark, above Opus 5.5's 59.
  • Neon's API pricing is $3 per million input tokens and $15 per million output tokens.
  • Google also opened its world model Genie 4 to AI Ultra subscribers the same day.
  • Genie 4 generates 720p video at 24 frames per second in real time with synchronized audio.

KEY DATA

61 pointsNeon composite intelligence score
59 pointsOpus 5.5 composite intelligence score
$3 per million tokensNeon input token price
720p/24fpsGenie 4 output quality

PHYSIX OBSERVATION

Google is rapidly deriving multiple models from the same underlying base, using dense releases to capture user mindshare, but Neon's unit price is higher than promotional-period Carbon, so its cost advantage holds only at the task level. For users, model selection will depend more on scenario than on version number; for the industry, the model iteration cadence is being compressed to weeks, sharply raising pressure on laggards.

Source: TMTPost report