DeepSeek 4.1 Flash Beats Gemini Flash: Faster, More Accurate, Far Cheaper
julianharris · x · 2026-09-19
The author spent two days evaluating models for content-quality assessment of software specs and found @typesafeai's small underlying model far too weak for the task. Update: DeepSeek 4.1 Flash is currently faster, more accurate and way cheaper than Gemini Flash. With a set of ground-truth data in hand, the author now plans to fine-tune a small model instead.
More from Models
- Jev Fails Probability Calibration: Says 99/1 When the Answer Is 60/40 — JnBrymn · 2026-09-19
- LLM fails coin-flip probability test, says 60/40 coin is 99/1 — JnBrymn · 2026-09-19
- Is Jev Actually Accurate? Engineer Flags 99/1 Answer to a 60/40 Coin Question — JnBrymn · 2026-09-19
- Pydantic Creator Benchmarks Jev vs Sonnet: 2.6x Cheaper, 3.7x Faster in 17 Lines — samuelcolvin · 2026-09-19
- Deep Dive: Open-weight models now within ~4 months of best closed frontier models — markjeffrey · 2026-09-19
- Veteran ML engineer: Jev may push agent tool-calling back to discriminative models — multiply_matrix · 2026-09-19