SaaS Feeds
Daily · 2026-10-17
26 reports from 10 outlets; 6 events saw new developments.
<built-in method items of dict object at 0x7d8010f56600> reports · 10 outlets · 6 events updated
Top events
Live board-
DeepSeek releases open-weight V4.1-Flash-1017 and opens its in-house search API
DeepSeek shipped V4.1-Flash-1017 as an open-weight checkpoint live on its website, app and API, reporting stronger agentic benchmark results including 77.0 on DeepSWE, faster task completion, shorter outputs and lower per-task costs than its predecessor. Alongside the release, DeepSeek opened up the in-house search API that powers its web and API search capabilities.
-
Meta releases Muse Spark Open, an 816B open-weight MoE model with 1M context
Meta released Muse Spark Open, an 816B-parameter open-weight mixture-of-experts model with 42B active parameters and a 1M-token context that leads open models on MCP-Atlas. The model is now available in Amazon SageMaker JumpStart and live on Together AI.
-
Anthropic's Claude solves a 25-year-old NIST cryptography puzzle in under 30 minutes
Anthropic's AI model Claude cracked a 25-year-old cryptography mystery in under 30 minutes, involving the mysterious seeds behind NIST's standard curves for digital signatures.
Agents and workflows
- Arena's Alignment Index: A New Metric for Agent Safety BestBlogs English AI
- Local AI Magic: Muse Spark Open on Mac Studio BestBlogs English AI
- Perplexity Decider Now Available on Vercel AI Gateway BestBlogs English AI
- Perplexity Open-Sources SPACE, a Firecracker MicroVM Sandbox for AI Agents MarkTechPost
- Nemotron 3.5 Super VL Multimodal Agent Model Technical Details BestBlogs English AI
- NVIDIA Nemotron 3.5 Bolt 24GB Local Agent Model Release BestBlogs English AI
- Building trust in AI agents: A production-ready evaluation framework with Amazon Bedrock AgentCore AWS Machine Learning
- Bottleneck in AI Editing: Rendering Speed BestBlogs English AI
Developer platforms
- Deepseek V4.1 Flash 1017 beats larger rivals in agentic tasks while once again massively cutting per-task costs The Decoder
- DeepSeek Just Released V4.1-Flash-1017, Scores 77.0 on DeepSWE, Opens Search API MarkTechPost
- StepFun Releases Step 5 Preview Open Weights: 600B MoE, 27B Active, 1M Context MarkTechPost
- Opening up the in-house search API DeepSeek API Updates
Enterprise
Product updates
- Automating Slide Deck Creation with Claude BestBlogs English AI
- Anthropics' Claude cracks a 25-year-old NSA cryptography mystery in under 30 minutes The Decoder
- Comparing Claude's Prompt, Project, and Skill Modes BestBlogs English AI
- NVIDIA's Midtraining Recipe for Agentic Coding: Boosting SWE Performance with Targeted Data Augmentation BestBlogs English AI
- Anthropic prepares to release Claude Fable 5.5 next week TestingCatalog
- Networking Opportunity: Hugging Face SF Party BestBlogs English AI
- v2.1.302 Claude Code Releases
- How Itaú is building generative AI for its entire customer base OpenAI News