llm-gemini 0.38: Gemini 4 Neon
Google introduced Gemini 4 Neon at $3/$15 per million tokens, currently the most expensive of the Gemini 4 models.
60 noviniek
Google introduced Gemini 4 Neon at $3/$15 per million tokens, currently the most expensive of the Gemini 4 models.
DeepSeek open-sourced DeepInfer, the production inference framework powering its large-scale services, supporting DeepSeek-V4.1-Flash with speculative decoding (DSpark), deterministic sampling, and structured outputs on NVIDIA SM100 and Huawei Ascend 950.
Google announced that Gemini in cars will gain access to users' personal context — calendar, home and work addresses, preferences — enabling more relevant navigation suggestions, messaging, and hands-free assistance while driving.
Anthropic updated the Developer Platform July 22-24 with API key expiration settings in the Claude Console with email reminders, Memory API ordering and stability fixes, and new Managed Agents features including effort levels and expanded webhook coverage.
The Model Context Protocol team released beta versions of Python, TypeScript, Go, and C# SDKs supporting the new stateless MCP 2026-07-28 spec. Teams building MCP servers or clients can test now before the final spec ships on July 28.
xAI launched a free Google Workspace add-on bringing Grok 4.5 to Google Docs, Sheets, and Slides, enabling formula building, chart creation, presentation drafting, and document editing powered by web and X data.
DeepSeek V4 graduated from preview to general availability with a 1M-token context window, V4-Pro and V4-Flash variants, MIT-licensed open weights, and the AI industry's first structural peak/off-peak API surge pricing.
Bolt.new introduced Skills -- reusable bundles of context, rules, and workflows (tech stack, design standards, code review rules) that Bolt applies during development, shareable across individual projects or entire workspaces.
Perplexity expanded its Agent API with GPT-5.6 Sol/Terra/Luna, Gemini 3.6 Flash, and Grok 4.5 at first-party pricing. Sonar Chat Completions is now consolidated into the unified Agent API.
Claude Code versions 2.1.215 through 2.1.218 (July 19–22, 2026) introduce manual-only /code-review and /verify invocation, new filesystem sandbox isolation control, emoji shortcode autocomplete in the prompt, and a fix for Windows path corruption with Unicode path segments.
Anthropic released a beta of Claude Security Plugin for Claude Code — a multi-agent tool that scans codebases for security vulnerabilities directly in the terminal during development, before commit. The plugin uses multiple sub-agents for parallel code examination.
Anthropic upgraded Claude's voice mode on July 23, 2026 — for the first time it runs on Opus and Sonnet models, not just Haiku. Paying users can now benefit from more capable, deeper-thinking models in voice conversations.
Anthropic launched Record a Skill in Claude Cowork: Pro, Max, and Team users can screen-record themselves doing a task and Claude turns the recording into a reusable skill. No prompt engineering needed — just show Claude how once.
xAI released a suite of free Microsoft 365 add-ins integrating Grok 4.5 into Excel, Outlook, Word, and PowerPoint. Grok in Excel helps write formulas and pivot tables; in Outlook it summarizes threads and drafts replies in the user's voice.
Google released Gemini 3.6 Flash with improved coding (+12pp on DeepSWE), the ultra-fast 3.5 Flash-Lite (350 tokens/s), and security-specialized 3.5 Flash Cyber — but the long-awaited 3.5 Pro is still absent.
Claude gained a monthly activity recap with AI observations (Settings > Reflect), Microsoft 365 write tools for drafting email and managing calendar events, and an Admin API for programmatic Enterprise organization management.
Perplexity Computer gains Brain — a private context graph built across sessions that boosted answer accuracy by 25%. New features include website publishing and full Microsoft 365 integration.
Google made gemini-omni-flash-preview available in Google AI Studio and the Gemini API — a natively multimodal video API priced at $0.10/sec of output with automatic SynthID watermarking. Also launched Nano Banana 2 Lite — Google's fastest and cheapest image model.
Stable Gemini CLI v0.51.0 (July 16, 2026) closes path traversal vulnerabilities, prevents model thought leakage into session history, and hardens the macOS sandbox. Nightly v0.52.0 (July 18) adds LLM triage orchestrator and ReAct loop mitigation.
Claude Code gets three key changes: artifacts now call MCP connectors live on each view (live dashboards), /fork creates a parallel background session without interrupting work, and a new EndConversation tool allows Claude to end abusive or jailbreak sessions.
xAI launched Grok Automations: users can schedule recurring AI jobs or trigger them via incoming email. Each execution starts a fresh Grok conversation, with results delivered via email, notification, or Slack.
Mistral AI released OCR 4, a document intelligence model with bounding boxes, block-type classification, per-word confidence scores, 170-language support, and a single-container self-hosting option for enterprises with strict data privacy requirements.
Google Vids now supports personal AI avatars (selfie + voice recording = SynthID-watermarked digital stand-in) and Gemini Omni Flash for text-prompt video generation and conversational editing of existing clips.
Vercel v0 now features a slash command menu for keyboard-first skill attachments, Shopify storefront integration with live product data, skill chips with draft persistence, and sound notifications on response completion.
Google renamed NotebookLM to Gemini Notebook and introduced new features: in-notebook code execution, integration into Google Search AI Mode and Gemini app, Gemini 3.5 model, and the Antigravity agentic coding platform.
DeepSeek introduced peak-hour pricing for its V4 API — token rates double during Beijing business hours (9–12 and 14–18 CST). The deepseek-chat alias retires July 24; migration to deepseek-v4-flash or deepseek-v4-pro is mandatory.
Claude Code released versions 2.1.207-211 with a new screen reader mode, subagent text streaming for JSON output, auto mode without opt-in on Bedrock/Vertex/Foundry, vimInsertModeRemaps, and a security fix for plugin hooks.
Anthropic launched self-serve HIPAA configuration for Claude Enterprise plan admins. Healthcare organizations can now enable HIPAA-compliant mode without contacting support.
Anthropic launched Claude for Teachers on July 14, 2026, giving verified US K-12 educators free access to Claude Opus 4.6, teaching skills grounded in learning science, and the Learning Commons connector with curricula mapped to academic standards in all 50 states.
Anthropic July 8-11 released configurable API key expiration with email notifications and Claude Code 2.1.206-207 with auto mode on Amazon Bedrock, Google Vertex AI, and IBM Foundry.
xAI released a major Grok Voice API update with 21 new flagship voices in 25+ languages, personal voice cloning from a 60-second sample, and expressive speech tags for developers.
DeepSeek will deactivate legacy aliases deepseek-chat and deepseek-reasoner on July 24, 2026 at 15:59 UTC. Developers must update calls to deepseek-v4-pro or deepseek-v4-flash — note critical mapping caveat for deepseek-reasoner.
The Gemini app gained Study Notebooks: a space to upload syllabi and notes, where Gemini diagnoses your knowledge level, creates adaptive lessons, and tracks progress across 100+ learning objectives. Launching on web for personal accounts.
Claude Code Desktop now includes a built-in browser panel (Cmd+Shift+B) where Claude can autonomously browse, click, and fill forms. The /doctor command now auto-fixes installation issues, CLAUDE.md bloat, and slow hooks.
Google significantly upgraded Gemini 3 Deep Think with iterative reasoning rounds for complex scientific problems. The model identified a logical flaw in a technical math paper missed by peer review. Available for Google AI Ultra subscribers.
Vercel AI Gateway adds ByteDance's Seedream 5.0 Pro — 2K output, layer editing, and reasoning-before-draw accessible through a single AI SDK endpoint without managing another provider.
Gemini CLI version 0.50.0 (July 8) introduces Tool Registry Discovery — automatic discovery and registration of external tools for AI agents.
Perplexity expanded the Sonar API with access to GPT-5.6 and Grok 4.5 models and added a new finance_search tool for real-time financial data.
Google, alongside the Gemini Spark expansion to the AI Pro tier (July 24), rolled out key improvements: the agent is 50% faster via parallel source retrieval and can now edit shared Workspace files.
Anthropic launched Claude Reflect in beta on July 9, 2026 — a 'Spotify Wrapped' for AI usage showing monthly topic breakdowns, peak activity patterns, and optional wellbeing settings like break reminders and quiet hours.
xAI released Grok 4.5 — its first model purpose-built for coding and agentic tasks, co-trained with Cursor on real engineering tasks. Available via xAI API at $2/M input tokens and $6/M output tokens, with a 500K-token context window and configurable reasoning effort (low/medium/high).
New Claude Code version sets Claude Opus 5 as default model, enables nested subagents to depth 3, brings stricter sandbox network controls, and improved MCP error reporting.
Anthropic expanded Claude Cowork with Microsoft 365 write access — Claude can now draft and send emails, manage calendar events, and create or update files in OneDrive and SharePoint on the user's behalf.
Anthropic launched Claude for Government Desktop in public beta with FedRAMP High authorization, bringing Claude Code, Cowork, hash-chained audit logs, and SCIM-based budget controls to U.S. government agencies.
SpaceXAI added /goal to Grok Build for handing off large implementation tasks autonomously, and /voice (Ctrl+Space) for hands-free prompt dictation — the first native voice interface in Grok Build.
Anthropic made Claude Cowork available on web and mobile on July 7—tasks run autonomously in the cloud even when devices are offline, with status notifications.
DeepSeek's official V4 release is coming mid-July with a 1-million-token context window and the platform's first-ever peak/off-peak API pricing — calls during Beijing business hours will cost 2x the off-peak rate.
Perplexity AI introduces a hybrid orchestrator that automatically routes AI tasks between local device and cloud frontier models based on data sensitivity and compute requirements — no manual configuration needed.
Anthropic released an expanded admin dashboard for Claude Enterprise: cost breakdowns by group and user, model-level entitlements, spend threshold alerts, and an Analytics API. Claude Code gets dedicated insight tabs updated daily.
Anthropic on July 22 updated the Claude Managed Agents API: configurable effort levels on model configuration, webhooks for environment and memory store lifecycle events, session seeding with initial events, and event deltas for subagent thread streaming.
xAI shipped a rapid series of Grok Build patches (0.2.66–0.2.73) over three days: sandbox profiles with kernel-deny lists, no-restart MCP server management, new agent dashboard with model visibility, and a fix for stdio hangs on Windows.
Claude Opus 4.8 and Haiku 4.5 are now GA in Microsoft Azure Foundry with native Azure authentication, consolidated billing through Microsoft Enterprise Agreements, and optional US data zone for compliance.
Anthropic released Claude Apps Gateway — a self-hosted control plane for enterprise Claude Code on Amazon Bedrock and Google Cloud, with SSO (OIDC), central policy enforcement, per-user spend limits, and OTLP telemetry.
xAI added 21 multilingual voices to Grok Voice on July 6 (25+ languages) and launched a no-code Voice Agent Builder for production voice agents with built-in telephony, MCP connectors, and observability.
The July 19, 2026 Claude Code update changes /verify and /code-review skill behavior — Claude no longer triggers them autonomously, only on direct invocation. Increases agentic pipeline predictability.
Anthropic raised Claude API rate limits across the board — Sonnet and Haiku now match Opus at every tier — and collapsed usage tiers into three (Start, Build, Scale), powered by a new SpaceX compute deal.
Vercel's AI Gateway adds beta support for realtime voice agents, text-to-speech, and speech-to-text, so developers can build conversational voice apps without juggling separate providers. Audio models route through AI SDK 7 with the same observability, BYOK and spend controls as text models, at no extra Gateway fee.
xAI's Grok audio family — realtime voice, TTS, and STT — is now routable through Vercel's AI Gateway with the same observability and spend controls as Grok's text models. Developers can compose voice agents with grok-voice-think-fast-1.0, grok-tts and grok-stt straight from AI SDK 7.
New `vercel metrics` command exposes Web Analytics directly via the CLI. Supports filters, dimensions, and is positioned for coding agents to answer questions about UTM campaign performance or mobile-vs-desktop conversion.
Concurrent build capacity for Pro teams jumped from 12 to 500. Enabled by default on Pro and Enterprise plans, billed on actual build minutes consumed.