Community cuts MiniMax H3 from 28 to 6 steps, topping the H3 Acceleration Arena
AdinaYakup · x · 2026-09-09
MiniMax's Ryan Lee shared open-source results for H3: released at 28 steps, the community then began optimizing it. After 11,850 votes in the H3 Acceleration Arena, several accelerated versions are statistically in the top group, with the current highest point estimate from @larryvrh's 6-step H3 Turbo v4 — a 28→6 step reduction driven by an ecosystem rather than one team, which the author cites as the payoff of open-weighting H3.
More from Models
- GPT-6 Astra Pro/Ultra review: first LLM to nail one-shot prompts nearly every time — rschu · 2026-09-09
- Frontier models ace counterexamples but lag on closed-form math, benchmark designer notes — evijit · 2026-09-09
- Raschka dissects GPT-6 Astra: looped transformers, recurrent depth, hidden CoT — rasbt · 2026-09-09
- Agentic video's real story isn't watching 90 minutes—it's reasoning about where to spend compute — romitheguru · 2026-09-09
- GPT-6 Astra system card draws fire: OpenAI claims 'most aligned model' ever — TheZvi · 2026-09-09
- Muse reportedly makes phone calls in Croatian, but users say it denies the ability — nathanbenaich · 2026-09-09