IBM has released Granite Switch 4.1 in 8B and 30B sizes, a modular system that runs chat, decision-making and speech transcription from one hosted endpoint instead of separate models. The update adds a general-purpose Decision adapter, a specialized Guardian adapter for safety-related classifications, and an embedded ASR model, Granite-Speech-5.0-Turbo-CTC, which tops the OpenASR Leaderboard. Despite bundling three capabilities, the combined model uses roughly the same VRAM as the base model alone.
Google has announced support for DPoP (Demonstrating Proof-of-Possession, RFC 9449) in its OAuth platform, cryptographically binding refresh tokens to a client-generated key pair so stolen tokens cannot be replayed by attackers. The protection applies only to long-lived refresh tokens, while short-lived access tokens remain standard Bearer tokens to avoid added cryptographic overhead on high-throughput API calls. Google also highlights DPoP's relevance for securing multi-agent and Model Context Protocol (MCP) architectures, recommending Google Cloud Secure Token Service for token delegation chains between agents.
Meta published Muse Spark Open, an open-weight mixture-of-experts model with about 816 billion total parameters and 42 billion active per token. It accepts text, images, audio and video as input and generates text, with a context window of up to 1 million tokens. Meta positions it as a coding-optimized, agentic model built on top of the smaller Muse Glimmer architecture, calling it the most capable US open-weights model released so far.
AWS published a governance framework called Trust in Depth, designed to let AI agents operate with more autonomy while staying within enforceable limits. It combines scoping agent permissions before deployment, giving each agent a verifiable identity, enforcing boundaries at runtime via policy gateways, and observing agent behavior afterward. The framework leans on automated reasoning (mathematical proof) for rules that can be precisely defined, and human judgment for ambiguous cases, using AWS tools like Bedrock AgentCore, Guardrails, Kiro, Strands, Agent Registry and CloudWatch Omni.
Swiss AI released Apertus 1.5, a continued pretraining of its open Apertus 1.0 models (8B and 70B, Apache 2.0) with native image and experimental audio understanding, a 4x larger context window of 262,144 tokens, an optional thinking mode, and improved instruction-following and tool use. The models were trained on a total of 19T (8B) and 17T (70B) tokens, with full training data, recipes and intermediate checkpoints published for reproducibility. The team also published detailed, standardized evaluation cards and announced early API access from providers like Swisscom and Infomaniak, plus a roadmap for a ~600B-parameter MoE Apertus 2.0.
Anam, which builds photorealistic interactive avatars, switched to Deepgram's Flux speech-to-text model to solve turn-taking problems in unscripted, real-time conversation. In A/B testing against a self-hosted open-source STT option, Flux cut unwanted avatar barge-ins by roughly 50% with no added latency penalty, and complaints about turn-taking on English sessions dropped to near zero. Anam deployed Flux self-hosted on its own Kubernetes/H100 infrastructure, rolling it out to 100% of turnkey users after a two-week self-hosted deployment (versus days for Deepgram's hosted option).
Researchers describe a system called FlowSpec that converts a creditor's legal call procedure into machine-readable instructions, tool restrictions, training signals and test cases for a voice agent. A 9-billion-parameter model was trained on calls covering nearly 300 randomly generated procedures, then tested on four procedures it had never encountered, including one entirely held-out family of scripts, to see whether compliance behavior transfers.
Google has released Apigee hybrid 1.17.0 (September 14, 2026), adding Model Context Protocol (MCP) support that runs entirely inside a customer's own Kubernetes cluster rather than on Google-managed infrastructure. This lets organizations with data-residency or regulatory requirements turn existing OpenAPI-described APIs into AI-agent-callable tools without MCP request traffic leaving their network. The feature reuses existing Apigee identity, security policies and API product controls, and it is an optional, off-by-default capability enabled via a Helm chart upgrade.
Anthropic rolled out a beta tracing feature that exports OpenTelemetry spans from both the Messages API and Claude Managed Agents sessions. Developers using the Claude SDK can now capture these spans natively in seven languages: Python, TypeScript, C#, Go, Java, PHP, and Ruby. Enabling the feature requires setting up a destination for the telemetry data inside the Claude Console.
xAI has rolled out Grok Voice 3.0, its latest speech-to-speech voice model, without changing pricing from the prior Grok Voice 2.0. Developers who reference the grok-voice-latest alias get the new model automatically, with no code changes needed. The API itself remains the same as before.
NVIDIA published a case study showing how Siemens EDA applied its open NeMo Gym evaluation library to tighten up an agentic tool used for chip characterization. Instead of checking individual tool calls, the team scored complete final output files with a domain-specific verifier, then used the failure patterns to rewrite the agent's skill instructions and MCP tool descriptions. After several rounds of this loop, a composite accuracy benchmark climbed from roughly 55% to 95%, while the agent needed far fewer steps and far less output to get there. The experiments relied on about 25 evaluation rollouts running Claude Opus 4.8 as the agent model inside Claude Code.
Cohere built a controlled testbed to study how teams of coding agents should be organized, testing where work is delegated, how leads wait on workers, which models play which role, team size, and communication patterns. On ProgramBench, well-organized teams outperformed single agents by up to ~10 percentage points but at 3–7x higher cost. The key finding: delegating only exploration and planning to workers matched the performance of delegating all stages (implementation and verification too) at half the cost, and a strong lead paired with 16 cheap planning workers reached 61.8% reward versus 63.2% for an all-strong-model team, cutting cost from $510 to $155.
GitHub replaced its LLM-based pipeline for detecting generic secrets (like passwords and API keys) with a fine-tuned ModernBERT classifier combined with regex candidate generation. The new two-stage system improved detection accuracy while cutting scan time and serving costs dramatically, and it's now fast enough to run during push protection, not just after code is committed.
Liquid AI has published LFM2.5-24B-A2B, a mixture-of-experts model with 24B total parameters but only 2B active per token, designed to run reasoning agents on local hardware. The release expands the context window to 131K tokens, doubles the vocabulary size, and was trained on roughly double the pretraining data of its predecessor. On agentic and tool-use benchmarks the model performs on par with or better than models that have about 50% more parameters, while delivering fast inference speeds on both consumer and data-center hardware.
Orca Security built Discovery AI, an Amazon Bedrock-based agent that turns plain-English security questions into structured queries against its Sonar system, which tracks over 900 cloud resource types. Rather than relying on one large model with the full schema stuffed into every prompt, the production version splits the work across nine specialized stages and multiple foundation models. The result: dramatically lower cost, faster responses, and better accuracy on complex queries, now running live since March 2026.
AWS detailed three kernel-level optimizations for Mixture-of-Experts (MoE) model serving on Hopper GPUs, implemented in collaboration with the vLLM community and shipped in Amazon SageMaker AI's LMI container. The changes target the inefficiencies of fused MoE computation during low-batch decode, which previously left Tensor Cores underutilized. Combined, the optimizations deliver up to 10% higher throughput and up to 9% faster time-to-first-token (TTFT).
NVIDIA has released the full Run:ai platform as open source, adding it as a foundation of the NVIDIA DSX AI Factory Platform for AI workload and GPU orchestration on Kubernetes. The release includes four modular components — KAI Scheduler, GPU Fractioning Operator, KAI Resource Management, and Workload Map — that can be adopted together or independently. Eight ecosystem partners, including Crusoe, Mirantis, Nebius, OpenNebula Systems, Rafay, Saturn Cloud, Spectro Cloud, and vCluster, are integrating the platform into their own offerings.
Apple ML researchers published a NeurIPS paper introducing SELFCOMPACT, a scaffold that lets language model agents decide for themselves when and how to compact their own accumulated context, rather than relying on fixed-interval summarization triggered by a token count. It combines a compaction tool the model can call with a lightweight rubric telling it when to fire (a sub-task is resolved) and when to hold off (mid-derivation or stuck). Tested on six benchmarks and seven models, it matched or beat fixed-interval summarization while using far fewer tokens.
Amazon Web Services researchers published a method called 'priming' that converts pretrained Transformer models into hybrid architectures mixing attention and state space model (SSM) layers, reusing existing weights instead of retraining from scratch. The approach cuts pretraining costs by over 99% and lets a 32-billion-parameter hybrid model match its source Transformer's quality while decoding up to twice as fast. Amazon tested three SSM types and released an open-source model zoo, training code, and serving tools under Apache 2.0.
Mistral AI has released Voxtral Mini 4B Realtime Arabic, a streaming speech-to-text model fine-tuned for Arabic dialects, Modern Standard Arabic, and code-switching between languages. The 4.4-billion-parameter model processes 16 kHz audio in real time with a configurable transcription delay, achieving an 8.82% average Character Error Rate across seven Arabic benchmarks at a 480ms delay — close to the offline Voxtral Transcribe Arabic model's 7.91% CER. It was co-developed with Morocco's digital transition ministry as part of a sovereign AI partnership and is released under Apache 2.0.
Anthropic has introduced a beta option on its Claude Platform that allows a Managed Agents agent to construct its own workflow for large, multi-part jobs, such as going through hundreds of documents. Instead of handling everything in one pass, the agent splits the task into phases, delegates pieces across multiple sub-agents, and merges their output into a single result. The job is handed off to the server, which executes it asynchronously and reports progress through a stream of events.
AWS published a joint engineering write-up with Postman describing how Agent Mode, Postman's AI-native interface for API testing and documentation, was built to run in production at scale. Rather than model quality, the team found that tool sprawl and missing context were the biggest obstacles to reliable agent behavior. The post details architectural patterns including dynamic tool scoping, schema-based query access, dedicated context handlers, and how Amazon Bedrock provides model flexibility, cross-Region inference, data residency controls, and prompt caching.
Google introduced AQuA (Ambient Quality Agent), a tool that runs continuously beside production AI agents inside a Google Cloud project to find failures that don't show up in logs or health checks, such as skipped validation steps or dropped context between sub-agents. It samples production session traces, grades them against a checklist, clusters recurring failure patterns, verifies clusters against full transcripts, and tracks them over time as new, recurring, or resolved. On request, it can also trace a failure back to the exact line of code or prompt responsible, using an immutable source snapshot captured at deploy time.
Google Cloud announced the Gemini agent, a unified agent that combines chat, knowledge work, content creation, and coding into one interface accessible across devices and channels. It integrates directly inside Gmail, Docs, Sheets, Slides, Chat, and Calendar, connects to enterprise tools and databases via MCP and a tools registry, and orchestrates across multiple models — including Google's Gemini family and Anthropic's Claude — to balance quality and cost. Google also detailed security/governance features, cost controls like Smart Routing and spend caps, and shared customer results from companies like BNP Paribas, Bradesco, and Orange Spain.