The pairing scores 68.1 on Artificial Analysis's Coding Agent Index v1.5 (DeepSWE 72.6, Terminal-Bench 4 62.1, SWE-Atlas-QnA 69.6) at $9.01 and about 55 minutes per task, effectively tied with Muse Code on Muse Wonder Full at 68.0. Cognition says that's 31% cheaper per task than Claude Code with Opus 5.5, a vendor claim, and you can try it now in Devin CLI and Desktop (Cognition).
Per The Information, Anthropic is testing a model that infers your real goal from thin instructions (one tester says it traces a single bug to the broader problem and offers a fuller fix), Google is readying Gemini 4 Carbon above last week's Argon, and OpenAI may ship a new Terra model within weeks that beats GPT 6.1 Sol on price. None of it is announced yet, so treat names and timing as unconfirmed…
Foundry agents now get a managed cloud browser (via Playwright Workspaces, provisioned as a separate Azure resource) that can reach public sites and private VNet apps, exposed to prompt or hosted agents through a toolbox's MCP compatible endpoint. Operators can watch sessions live, take over mid run, and require approval before clicks, typing, or uploads, which you enable only when a task needs them…
On OpenAI's new MultiLangBench, where a model builds a SQLite compatible engine in 20 languages, GPT 6 Astra averages an 87.4% test pass rate (COBOL jumps from 10.9% to 72.1%) versus 26.8% for GPT 5.6 Sol. On its COBOL migration eval, GPT 6.1 Sol hits 86.6% at low reasoning effort and 97.9% at high, though both benchmarks are OpenAI's own and not independently verified yet…
Built on MCP, the connectors let an agent carry engineering work from CAD (CATIA, NX) through meshing (ANSA, HyperMesh) to simulation (Abaqus) across Windows workstations and Linux clusters, with engineers setting goals and validating results; one customer cut a wheel rim geometry cleanup from a day to about three minutes. It's available now, but only to existing Neural Concept customers (SiliconANGLE)…
The filesystem, memory, sequential thinking, and everything servers move to TypeScript SDK v2 and the 2026 07 28 spec while still answering older clients, and the Python fetch, git, and time servers ( 2026.10.19 ) now return a JSON RPC 32602 error for unknown tools instead of an isError result. The TypeScript servers require Node 24 or later, and v1.0.0 (Oct 11) was the final legacy era release, so pin it if you can't upgrade yet…
Spotted something we missed? Start a thread.