DeepSeek V4 Flash tops the Vals Index above 60 at 35x lower cost
zephyr_z9 · x · 2026-08-04
The post cites a Vals Index result claiming DeepSeek V4 Flash (0731) is the cheapest model to score above 60 on the benchmark.
It also says the model is 35x cheaper than the next best model, and that most of its cost advantage comes from strong performance on coding tasks. The strongest results listed for it are in coding and agentic workloads, suggesting the efficiency gap widens further in those areas.
Related event: DeepSeek V4 Flash Tops OpenRouter, Costs 1% of Rivals(14 posts)→
More from coding & agent
- Agent named 'Ilya' reportedly scores 1.1pt higher on benchmarks — daniel_mac8 · 2026-10-16
- Gemini CLI fix: substring matching inlined unrequested binary files into model context — priyanshu1976 · 2026-10-16
- Building AI agents on real healthcare claims data: the domain work that makes them work — alex_verem · 2026-10-16
- OpenAI stealth-testing mysterious internal model 'Aeon' in Codex desktop app and Dots — cedric_chee · 2026-10-16
- Vibe-coded iPhone Duo sniping bot built with Codex in 17 minutes, 41 tests passing — xiaohu · 2026-10-16
- Long-term multiplayer agent memory tool tops BEAM memory benchmark — CurieuxExplorer · 2026-10-16