FlappyBench Showdown: DeepSeek V4.1 Flash and Qwen Omni Flash Beat Gemini 3.8 Flash on Cost and Quality

Command Code 6 hours ago

Key Info

A developer-run FlappyBench test compared three AI coding models building a Flappy Bird-style game from the same design prompt. Qwen 3.8 Omni Flash and DeepSeek V4.1 Flash both scored 9/10, with DeepSeek delivering a one-shot solution at the lowest cost, while Gemini 3.8 Flash scored only 5/10 after multiple iterations and cost roughly 23x more.

Highlights

Qwen 3.8 Omni Flash: 9/10 at $0.013 with smooth gameplay.
DeepSeek V4.1 Flash: 9/10 at $0.0089, solved in a single pass — the cheapest result.
Gemini 3.8 Flash: 5/10 at $0.205, needed 4–5 iterations, still produced hard-coded gameplay.

Original post

Ran FlappyBench on Qwen3.8-Omni-Flash, DeepSeek-V4.1-Flash, and Gemini 3.8 Flash with the same design prompt. 🔹 Qwen 3.8 Omni Flash: 9/10 · $0.013 · smooth gameplay and cheaper 🔹 DS V4.1 Flash: 9/10 · $0.0089 one-shot with the lowest cost 🔹 Gemini 3.8 Flash: 5/10 · $0.205 · 4-5 iterations but still hard-coded gameplay and costs 23x more

#FlappyBench#Qwen 3.8 Omni Flash#DeepSeek V4.1 Flash

View original on X