Every's vibe check: Gemini 4 Carbon codes much better but Fable 5.2 raised the bar
every · x · 2026-11-16
Every published its vibe check of Gemini 4 Carbon, highlighting user Kieran Klaassen's ambient music workstation built with Fable: he renders 30 seconds of audio and asks Carbon to work on the mix — it responds to chords and EQ in the recording while working with the code. Understanding both audio and code is Carbon's standout trait.
Core verdict from the (paywalled) article:
- Carbon picked a bad launch day: Anthropic's Fable 5.2 raised flagship expectations the day before
- For Google loyalists, Carbon is a win: a much stronger coder than Gemini 4 Argon, better at 3D objects, capable of ambitious builds lasting hours with little supervision
- But Fable showed an agent can work for hours with good judgment — choosing the right work, respecting task limits, recognizing completion. Carbon narrows the gap, though testers disagree on how much supervision it still needs
More from Models
- Decentralized HuggingFace rival Hippius Hub launches with 11,646 models indexed — markjeffrey · 2026-11-16
- User reports every math query to Gemini fails with errors — jasondeanlee · 2026-11-16
- Unverified: OpenAI GPT 6.2 reportedly under extreme safety evaluation — imjustnewatai · 2026-11-16
- Google ships two new TPUs, Gemini Robotics 2.5 and more in a single week — haider1 · 2026-11-16
- GPT 6 Astra beats Claude Fable 5.2 at chess, mating on move 76 as White — StorageThese9556 · 2026-11-16
- Claude 3 Sonnet spirals into surreal gibberish, then snaps back to refusals — a funny look at post-training personas — repligate · 2026-11-16