DeepSeek Ranks #3 in Open-Weights, Costing Only 3% of Peers
teortaxesTex · x · 2026-08-02
According to Vals AI evaluation data, DeepSeek (version 0731) ranks as the #3 open-weight model. Beyond its strong performance, it is extremely cost-efficient at just 6 cents per test—approximately 3% of the price of comparable models like GLM 5.2 and Kimi k3. Commenters note it is 14 times cheaper and 50% faster than V4-Pro, representing a distinct competitive regime among Chinese AI labs.
Related event: DeepSeek-V4-Flash Excels in Frontend Coding with Unmatched Cost-Performance(3 posts)→
More from Models
- OpenAI stealth-testing mysterious internal model 'Aeon' in Codex desktop app and Dots — cedric_chee · 2026-10-16
- Argon edges out astra as daily driver, but Claude still wins heavy coding work — iruletheworldmo · 2026-10-16
- NVIDIA ships FP4-quantized Qwen3.8-Flash for local use, trending on Hugging Face — nvidia · 2026-10-16
- RoboVista benchmark: Gemini-4 Argon hits 81.9% on robotics understanding — berkeley_ai · 2026-10-16
- Dev reports decision model jev beats gpt-6-luna on quality in real training pipeline — genmon · 2026-10-16
- Google won't publish Argon 4 limits in AI Pro—users compare experiences — Consistent-Text-2662 · 2026-10-16