aiminute. ← All AI news
New Models AI Minute Newsroom 2026-11-19

Google's mid-tier Gemini 4 Neon edges past its own flagship in tests.

Google's mid-tier Gemini 4 Neon edges past its own flagship in tests.

Google released Gemini 4 Neon on Wednesday for paying users of its API, Gemini app and Antigravity coding tool. Independent tester Artificial Analysis scored it 61, one point above Google's flagship Carbon and third overall. It produced output about twice as fast as Carbon and finished test tasks in roughly 21 minutes versus 36.

Why it mattersNeon costs $3 per million input tokens, which is above Carbon's discounted price until November 30. If you pay for Gemini, your default model just got faster without getting weaker.

✓ Verified · 3 sources

WhatsApp X Telegram
Read in the app — free, in 9 languages

Related stories

Kuaishou's new Kling image model generates pictures in native 8K.
2026-11-19
Microsoft unveils its own coding model to undercut rivals on cost.
2026-11-18
OpenAI's cheaper new Astra model still trails Claude in tests.
2026-11-18
French cloud firm OVH unveils its first language model, Olala mini.
2026-11-17
Tencent releases Hy4 flash as a cheaper, faster version of its model
2026-11-16