Gemini 4 Neon feels better-judged than Carbon, but cache hit rates and TTFT are bad
lyraxana · x · 2026-11-20
A reposted hands-on impression: Gemini 4 Neon is described as a great, pleasant-to-work-with model with better judgment than Carbon in the author's experience. Two engineering pain points hold it back — terrible cache hit rates and bad TTFT (time-to-first-token); fixing those would make the experience even better. Limited detail, no numbers, but a rare third-party comparison of these two models.
More from Models
- LMArena publishes Peter Gostev's leaked system prompts collection on YouTube — arena · 2026-11-20
- Cloudflare's clairic, a Qwen3.5-based multimodal info-extraction model, trends on Hugging Face — Cloudflare · 2026-11-20
- Google's comeback draws 'this is a real Code Red' reaction as rivals feel the heat — iruletheworldmo · 2026-11-20
- Google's Live Interactions API goes live: Genie 4 gets native WebRTC and WebSocket support — _philschmid · 2026-11-20
- Google abruptly removes Argon persona mid-conversation, upsetting devoted users — amplifiedamp · 2026-11-20
- Opinion: One Cheap Chinese Open-Weights Model at Parity with Opus 5.5 Means Game Over for Anthropic — IridiumEagle · 2026-11-20