DeepSeek V4 Flash Matches GLM 5.2, Tops OpenRouter Rankings
natolambert · x · 2026-08-05
Nathan Lambert points out that the adoption numbers for DeepSeek V4 Flash have been insane, though underrated in public discourse.
- Benchmark Performance: Epoch AI reports an ECI score of 153, matching GLM 5.2 and sitting between Opus 4.5 and Opus 4.6, making it the second strongest open-weights model today (behind Kimi K3).
- Ecosystem Adoption: Lambert calls the new version a "total monster" poised for extensive use. It is already the top model on OpenRouter with massive activity on HuggingFace.
Related event: DeepSeek V4 Flash Tops OpenRouter, Costs 1% of Rivals(14 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