Mollick: Gemini 4 is first non-OpenAI/Anthropic model to pass the Fable/Mythos/Astra bar
emollick · x · 2026-10-14
Ethan Mollick spent extensive time with Gemini 4 pre-launch: it's a big jump over previous Geminis and the first non-OpenAI/Anthropic model to pass the Fable/Mythos/Astra bar. It's more pleasant and agentic than the Gemini 3 series, with a distinct personality from Claude or Astra. Creatively it matches OpenAI and Anthropic, but lags at complex multi-agent work, tends to be overconfident and a bit lazy — issues amplified by the Antigravity harness trailing competitors, though he expects rapid improvement.
More from Models
- Argon model 'incredibly cooked': skips reasoning and leaks its chain-of-thought as plain text — scaling01 · 2026-10-14
- Fields Medalist Gowers says Grok solved a graph theory problem he studied for a year — paul_cal · 2026-10-14
- LiquidAI's edge-focused MoE LFM2.5-24B-A2B trends on Hugging Face — LiquidAI · 2026-10-14
- Gemini 4 Argon goes GA in GitHub Copilot, beating prior Gemini models on coding — 0xkarasy · 2026-10-14
- Gemini 4 Argon lands in Cursor: 52.8% on CursorBench, matching Opus 5.5 at $2/$10 — gaganghotra_ · 2026-10-14
- User reports Claude is far less deceptive, nearing Codex-level attention to detail — DevDminGod · 2026-10-14