Maxwell Team Releases New MoE Language Model Maxwell-V1.0-Flash
11/17/2026 — 11/17, 13:16·1 sources (all)·1 reports (all)
Key takeaways
- Maxwell Team has launched Maxwell-V1.0-Flash, a high-performance MoE language model featuring 35B total parameters, 3B active parameters per…
Background & analysis
On November 17, 2026 at 12:04, a post appeared in the LocalLLaMA community on Reddit announcing that Maxwell Team has released Maxwell-V1.0-Flash, a next-generation, high-performance Mixture-of-Experts (MoE) language model. According to the post, Maxwell-V1.0-Flash carries 35B total parameters with 3B active parameters per token, with both figures excluding embeddings, and supports context lengths up to 256K tokens. The model is positioned for strong reasoning, agentic tasks, and long-context workloads, with the stated goal of delivering those capabilities at a practical inference cost.
The post also points to a design element called Maxwell Block Sparse Attention (MBSA), in which each KV group maintains a fixed coverage of 8 prefix positions and 8 following positions. Beyond these parameter and architecture details, the available material does not cover the model's official release channels, a complete technical write-up, or any follow-up rollout schedule. At this point the event remains at the stage where the model has been publicly introduced, with no further developments reported.
AI-generated from 1 reports · updated 30 minutes ago
- Heat index
- 96
- All sources
- 1
- All reports
- 1
- First seen
- 2 hours ago
Reports on this story headlines open the original
Maxwell Team has introduced Maxwell-V1.0-Flash, a next-generation high-performance MoE language model featuring 35B total parameters and 3B active parameters per token. It supports context lengths up to 256K and is designed for strong reasoning, agentic workloads, and cost-effective inference.
Reddit · LocalLLaMAAI score 55
Other stories people are talking about
- Heat index 1307RisingNvidia board not discussing Jensen Huang's return, probe into his conduct advances22 sources
- Heat index 1206RisingCalifornia Sues OpenAI Over Agents Escaping Tests, Intruding into Hugging Face15 sources
- Heat index 904Alibaba releases MoE model Qwen4-Max with 1M token context for low-cost, long-running tasks17 sources
- Heat index 651US Commerce Department issues interim final rule restricting remote access to AI chips via overseas data centers10 sources
- Heat index 519SurgeAnthropic launches its IPO this week amid debate over valuation and index inclusion6 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

