Poolside releases Laguna S 2.1, a 118B open model with 1M-token context
willccbb · x · 2026-07-22
Poolside released Laguna S 2.1, its most capable model so far.
Key details:
- 118B total parameters, 8B active per token
- Up to 1M token context window
- Supports both thinking and non-thinking modes
- The team says it can compete with models many times larger while still fitting on a single NVIDIA DGX Spark
- Weights are available now under OpenMDW-1.1 on Hugging Face
The image also shows benchmark comparisons where Laguna S 2.1 leads several coding and multilingual evaluation suites, including Terminal-Bench, SWE-Bench Multilingual, SWE-Bench Pro, DeepSWE, SWE Atlas, and Toolathlon Verified.
Related event: Poolside Releases 118B Open-Weight Coding Model Laguna S 2.1(35 posts)→
More from Infra
- DeepSpeed shows AutoEP/AutoTP/AutoSP parallelism gains, hosts office hours — PyTorch · 2026-10-27
- Cloudflare Durable Objects validated as top architecture for AI agent WebSocket servers — tristanbob · 2026-10-27
- Gary Marcus: With Q3 data added, Anthropic's compute-efficiency growth shows deceleration — GaryMarcus · 2026-10-27
- Inference is just Kubernetes with a multi-trillion-dollar market, and the communities overlap — matt_slotnick · 2026-10-27
- vLLM eyes Rust migration as inference engines race to rewrite the stack — AccBalanced · 2026-10-27
- Resin lets an agent build standalone LLM inference engines; vLLM and SGLang optimize DeepSeek kernels differently — AccBalanced · 2026-10-27