Tuesday Jul 7
RESEARCHOther

Tencent Ships Hy3, Free For Two Weeks

6JUL
HY3

China just gave away a model that rivals the world's best. Tencent open-sourced Hy3, a 295-billion-parameter model, free through July 21. It only wakes up a small slice of itself for every answer.

Tencent just released one of the most efficient open models yet. It's called Hy3, and anyone can download or rent it free.

Hy3 has 295 billion total parameters but only turns on 21 billion per answer. That keeps compute costs low while chasing GPT and Claude-level scores. It hits 90.4 on GPQA Diamond, a hard science benchmark.

The weights are free under Apache 2.0, no strings attached. It's the latest sign China's open models are closing the gap fast.

full brief & sources

⚡ Why this matters

  • Open-weight models rivaling flagship performance change the buy-vs-build math for every AI product team.
  • China's labs are shipping efficient models faster than US open-weight competitors right now.
  • Free access on OpenRouter means any developer can test it today, no waitlist.

🔍 What happened

  • Jul 6: Tencent Hunyuan released Hy3, a 295B mixture-of-experts model.
  • Only 21B of the 295B parameters activate per token, 192 experts with top-8 routing.
  • 256K context window, Apache 2.0 license, weights on Hugging Face.
  • Free on OpenRouter through July 21.
  • Scores 90.4 on GPQA Diamond, 72.0 on USAMO 2026.
  • Built for reasoning, agentic workflows, and long-context tasks.

💬 Smart takes

  • Simon Willison: flagged Hy3 within a day of release as worth watching.
  • MarkTechPost: Hy3 approaches flagship-model performance at a fraction of the active-parameter cost.
  • Skeptic: benchmark scores on launch day rarely survive contact with messy real-world prompts.

🧭 Where this goes

  1. LikelyHy3 gets adopted fast by cost-sensitive startups once the free OpenRouter window ends.
  2. LikelyUS labs respond with their own efficiency-focused open releases within the quarter.
  3. PossibleHy3's sparse routing approach gets copied by the next wave of open models.
  4. Wild Carda major enterprise standardizes on Hy3 over a US flagship model for cost reasons.

🥄 The Spoon Take

The efficiency race matters more than the size race now. Hy3 proves you don't need all 295 billion parameters awake to compete with the frontier. Every lab chasing bigger models should be nervous about labs chasing cheaper ones instead.

🤔 Pushback

Free-for-two-weeks pricing is a promotion, not a business model. The real test is what Hy3 costs once the trial ends.

·MarkTechPost - Tencent Hy3 295B release·Simon Willison - tencent/Hy3 notes·WinBuzzer - Hy3 benchmark rundown