JetBrains open-sources Mellum2.1, a small coding agent model
JetBrains has open-sourced Mellum2.1, a 12B MoE model with 2.5B active parameters. It scores 47.0 on SWE-bench Verified, up from 2.0 for its predecessor.
Agent News and Analysis
13 verified stories covering Agent, product updates and industry developments.
Poolside Releases 118B Open Coding Model
Laguna S 2.1 gives enterprises a self-hosted coding-agent base, trading frontier scores for local control and clearer deployment economics.
Agnes-2.5-Flash uses a free coding model to pressure agentic developer tools
Agnes-2.5-Flash uses a free coding model to pressure agentic developer tools. The piece reviews the verified facts and why the signal matters beyond one announcement.
Alibaba tests agents with HS code classification
Alibaba tests agents with HS code classification is reframed for global technology readers, with the key numbers, product claims and open questions kept intact.
Fake Sentry errors can hijack coding agents
Tenet Security describes Agentjacking, where attackers inject fake error reports that coding agents treat as trusted repair instructions, exposing secrets and deployment credentials.
OpenAI acquires Ona to deepen Codex infrastructure
OpenAI’s purchase of the Gitpod-linked cloud sandbox company points Codex toward managed development environments, not just code suggestions.
MiniMax M3 brings sparse attention, supports million-token context
MiniMax released the open-weight M3 model with 1M-token context, advanced coding ability, and native multimodality, claiming it is the first to combine all three.
Google’s Gemini Spark costs $100/month, connects 11 SaaS platforms
Google I/O 2026 introduces Gemini Spark, a 24/7 AI agent with its own Gmail address, priced at $100/month and integrated with 11 SaaS services via MCP.
Gemini 3.5 Flash beats GPT-5.5 in agentic benchmarks, costs just $1.5
Google's Gemini 3.5 Flash outperforms GPT-5.5 on agentic benchmarks while being dramatically cheaper, and becomes the default model across Gemini app and Search.
PwC to train 30,000 consultants on Claude as Anthropic targets $2 trillion tech debt
PwC and Anthropic expand their partnership, with 30,000 PwC consultants learning Claude to tackle an estimated $2 trillion in global enterprise tech debt.
EY equips 130,000 auditors with AI Agent
EY has deployed enterprise-scale agentic AI globally, covering 130,000 audit professionals across 160,000 engagements in 150 countries. The multi-agent system handles complex task coordination, dynamic risk processing, and administrative burden reduction, with full end-to-end AI-supported audits expected by 2028.
Canva shifts to AI platform with 265 million monthly active users and $42 billion valuation
Canva unveiled its "biggest product upgrade ever" at Canva Create 2026, repositioning itself from a design tool with AI features to an AI platform with design capabilities, introducing conversational design, persistent memory, automated scheduling, and MCP connectors.
IT Leaders Say Enterprise AI Agents Are Sprawling Fast
OutSystems survey data shows broad AI agent adoption, but governance, audit trails and EU AI Act readiness are lagging badly.