Zai releases GLM-5.3 Flash: 320B params, 2x efficiency on DeepSWE
togethercompute · x · 2026-08-28
Zai has released GLM-5.3 Flash, a natively multimodal model with 320B total parameters, 18B active, 1M context, and hybrid attention. On the DeepSWE benchmark, it nearly matches Luna's performance while completing more than twice the work for the same budget.
Related event: Zai Open-Sources GLM-5.3 Flash, a 320B Native Multimodal Model(3 posts)→
More from Models
- Google's Gemini Omni 1.1 Flash takes #1 in Text-to-Video Arena, #2 in Image-to-Video — sedielem · 2026-08-28
- User Returns to ChatGPT to Escape Claude's Weird Vocabulary — evielync · 2026-08-28
- Introducing PhoneLLM: Open Source Voice Model with 1/3 Latency and 1/18 Cost of GPT-5.6 Terra — charles_irl · 2026-08-28
- Omni 1.1 Flash Released on LMSYS Chatbot Arena — PleasantSir9581 · 2026-08-28
- Quality Difference: Minimax bf16 vs Pruned/int8 — Dapper_Astronaut_603 · 2026-08-28
- Cartesia's Sonic-3.6 tops TTS leaderboard, beats ElevenLabs in real-world test — socialwithaayan · 2026-08-28