Speculation on Kimi K3.1 and DeepSeek V4 performance and pricing
Linahuaa · x · 2026-07-20
The author speculates that a more polished Kimi K3.1 could reach the level of Fable/Sol or slightly above, while using tokens more efficiently.
They also predict the upcoming DeepSeek V4 could land between Opus 4.8 and Kimi K3 in capability, while being dramatically cheaper. The post ends with a taunt that Western labs will need to ship something exceptional soon or AI Twitter will go “full commie.”
More from Models
- Anthropic says Opus 5.5 cuts cost per task by ~40% vs Opus 5 — taufiqintech · 2026-10-26
- AI search APIs ranked: top 12 within 7 points, yet prices differ nearly 7x — FinanceYF5 · 2026-10-26
- OpenAI Dots coming to EU/EEA and UK in a few weeks, slides confirm — testingcatalog · 2026-10-26
- OpenAI Dots demoed live as EU launch speculated, Codex adoption in Germany revealed — testingcatalog · 2026-10-26
- "Abliterated models just have brain damage" — community pushes back on unaligned model hype — dhadfieldmenell · 2026-10-26
- dair-ai publishes primer on Cloudflare's Clairic System One extraction model — dair_ai · 2026-10-26