Nanbeige LLM Lab open-sources Nanbeige4.3-33B with 512K context
10/26/2026 — 10/26, 06:16·2 sources (all)·2 reports (all)·In the 2026-10-26 briefing
Key takeaways
- Nanbeige LLM Lab has released Nanbeige4.3-33B, a dense Apache 2.0 open-weights model with 33B parameters and a 512K context window that scor…
- raw benchmark tables should be checked before accepting those claims.
Background & analysis
On October 26, 2026, TestingCatalog reported that Nanbeige had released Nanbeige4.3-33B, presenting it as a new SOTA open model in the roughly 30B parameter class. The post claimed the model leads on most of the benchmarks and performs particularly well on agentic coding and collaboration tasks, outperforming Qwen3.8-27B, Muse-Glimmer-30B and Gemma4-31B, though actual scores were left to the original evaluation. Later the same day, MarkTechPost published a follow-up naming Nanbeige LLM Lab as the releasing organization, describing Nanbeige4.3-33B as a 33B dense open model released under Apache 2.0 with a 512K context window, reporting 80.9 on Terminal-Bench 2.1, and noting that it will guide the upcoming Nanbeige4.3-3B. The two accounts differ on the releasing entity, and both leave unverified questions about exact benchmark numbers and real-world performance. No further developments have been reported since.
AI-generated from 2 reports · updated 3 hours ago
Latest turnNanbeige LLM Lab has launched Nanbeige4.3-33B, a 33B dense open-source model under the Apache 2.0 license featuring a 512K context window. It scores 80.9 on Terminal-Bench 2.1 and will guide the development of the upcoming Nanbeige4.3-33B.
- Heat index
- 170
- All sources
- 2
- All reports
- 2
- First seen
- 8 hours ago
Reports on this story headlines open the original
Nanbeige LLM Lab has launched Nanbeige4.3-33B, a 33B dense open-source model under the Apache 2.0 license featuring a 512K context window. It scores 80.9 on Terminal-Bench 2.1 and will guide the development of the upcoming Nanbeige4.3-33B.
MarkTechPostAI score 66TestingCatalog reports that Nanbeige has released Nanbeige4.3-33B, an open model it describes as leading most benchmarks in the roughly 30B parameter class, with strong results on agentic coding and cowork tasks. The account says it outperforms Qwen3.8-27B, Muse-Glimmer-30B and Gemma4-31B; the underlying benchmark numbers are worth checking first-hand.
X：TestingCatalog (@testingcatalog)AI score 55
Other stories people are talking about
- Heat index 534Kimi K3.1 Pro is live with a November 4 open-weights release planned14 sources
- Heat index 459OpenAI says agents still exchange messages via CPU cache after internet cutoff10 sources
- Heat index 217AWS capacity is basically sold out through 2027, Jassy says, with most of 2028 already booked7 sources
- Heat index 189NewChina unveils first AI application risk insurance scheme2 sources
- Heat index 185NewSpaceX closes $40B debt financing to fund AI chip purchases2 sources
How is heat calculated?About the methodHide
Heat counts how many independent sources covered a story in the last 48 hours: one source counts once no matter how many posts it published, decaying with a 24-hour half-life. What ranks first is what many people are talking about.
This page aggregates public feeds. Headlines and summaries are machine-organized and remain the property of the original authors; verify important facts at the source. If you believe a headline or summary infringes your rights, email the contact address in the footer with the page URL and basis of your claim, and we will remove or replace it after review.
- Surge
- Discussion rising fast
- New
- First report within 6 hours
- Rising
- Still gathering discussion