Looking Back at the DeepSeek V3 and R1 Impact
kevinnbass · x · 2026-07-14
This post reviews the industry shockwaves caused by DeepSeek V3 and R1 at the time:
- The author considers R1 to be DeepSeek's "breakout moment," released about a month after V3.
- V3 was also highly capable, approaching GPT-4o levels at a fraction of the cost, but it didn't generate the massive attention that R1 did.
- Because R1 could be directly compared to o1, it made a wider audience realize the disruptive power of this reasoning model approach.
- The author adds that their own experience using R1 outside of chat scenarios wasn't great, as it often fell into weird loops.
- The comments section reflects further: during past debates over vibe coding, many engineers argued that LLMs could never create anything new and would be limited by training data; however, looking back today, the models from that era pale in comparison to what we have now.
Related event: New Model Generations Bring AGI Closer to Reality(3 posts)→
More from Models
- Sam Altman is headed to Washington to brief Congress on OpenAI’s GPT-6 line — inductionheads · 2026-07-22
- Benchmark chart pits GPT-5.6 Luna, Grok 4.5 and Gemini 3.6 Flash on price and scores — iruletheworldmo · 2026-07-22
- Claim says Kimi was distilled from Fable, sparking a model-attribution jab — cephaloform · 2026-07-22
- Gemini 3.6 Flash is now available in Antigravity and chat — MartianOnJupiter · 2026-07-22
- Gary Marcus says LLM math skills are like knowing only a car’s engine size — GaryMarcus · 2026-07-22
- OpenAI’s Codex + GPT-5.6 Sol hits 99% recall in Project APE verification tests — soumitrashukla9 · 2026-07-22