The Deploy Log | Month by month
58 rows dated April 2026 are in The Deploy Log: 11 leads, 16 in Models and APIs, 25 in Tools and products, 2 in Robotics, hardware and chips and 4 in Breakthroughs, from 9 publishers. They include ChatGPT for Clinicians from OpenAI, GPT-5.5 from OpenAI, Workspace agents from OpenAI, Claude Design from Anthropic, Claude Opus 4.7 from Anthropic and 53 others. Every item sourced to the publisher's own page.
ChatGPT for Clinicians is now available for free to verified physicians, NPs, PAs, and pharmacists in the U.S. It includes frontier models, trusted clinical search, deep research across medical journals, and CME credit from real clinical questions.
GPT-5.5 and GPT-5.5 Pro are now available in the API, rolling out to Plus, Pro, Business, and Enterprise users in ChatGPT and Codex. Terminal-Bench 2.0 hits 82.7%, GDPval 84.9%, OSWorld-Verified 78.7%, and Tau2-bench Telecom 98.0% without prompt tuning.
Workspace agents are now generally available in ChatGPT Business, Enterprise, and Edu. They run in the cloud, keep working when you are away, and can be deployed in Slack. Free until May 6, 2026, then credit-based pricing.
Claude Design is available in research preview for Claude Pro, Max, Team, and Enterprise subscribers. It is powered by Claude Opus 4.7 and exports to Canva, PDF, PPTX, or standalone HTML files.
Claude Opus 4.7 is generally available at $5 per million input tokens and $25 per million output tokens. On CursorBench it clears 70% versus Opus 4.6 at 58%, and on Rakuten-SWE-Bench it resolves 3x more production tasks than Opus 4.6.
GPT-Rosalind is available globally to eligible organizations through OpenAI's trusted-access program. On LABBench2 it outperforms GPT-5.4 on 6 out of 11 tasks, and best-of-ten submissions ranked above the 95th percentile of human experts on an RNA sequence-to-function prediction task.
Waymo opened its fully autonomous ride-hailing service to public riders in Nashville with an initial 60-square-mile service area covering Broadway, 12 South, Midtown, and East Nashville.
Overworld released Waypoint-1.5, a real-time video world model with 720p and 360p tiers that runs locally on RTX 3090 through 5090 hardware at up to 60 FPS.
Teams on ChatGPT Business and Enterprise can add Codex-only seats with pay-as-you-go pricing and no rate limits. The annual price of ChatGPT Business drops from $25 to $20 per seat. Eligible workspaces receive $100 in credits per new Codex-only team member, up to $500 per team.
Gemma 4 ships in four sizes: Effective 2B, Effective 4B, 26B MoE, and 31B Dense. The 31B ranks #3 among open models on the Arena AI text leaderboard, and the 26B ranks #6. The 26B MoE activates only 3.8 billion parameters during inference.
OpenAI closed a funding round with $122 billion in committed capital at a post money valuation of $852 billion. The company is generating $2 billion in revenue per month, and ChatGPT has more than 900 million weekly active users and over 50 million subscribers.
OpenAI introduced Advanced Account Security, an opt-in setting requiring passkeys or security keys, disabling email/SMS recovery, and shortening sessions, with Yubico bundle pricing.
DeepInfra is now a supported Inference Provider on Hugging Face Hub, offering over 100 models including DeepSeek V4, Kimi-K2.6, GLM-5.1, with PRO users getting $2 monthly Inference credits.
IBM released Granite 4.1, a family of dense LLMs (3B, 8B, 30B) trained on ~15T tokens with 512K context, Apache 2.0, with the 8B matching the previous 32B MoE.
DeepSeek released V4 today with two MoE checkpoints on the Hub: DeepSeek-V4-Pro at 1.6T total parameters with 49B active, and DeepSeek-V4-Flash at 284B total with 13B active, both with a 1M-token context window.
The Agents SDK now includes native sandbox execution with built-in support for Blaxel, Cloudflare, Daytona, E2B, Modal, Runloop, and Vercel, plus snapshotting and rehydration so losing a sandbox container does not mean losing the run.
Cloudflare is expanding access to OpenAI frontier models including GPT-5.4 across Agent Cloud, and the Codex harness is now generally available in Cloudflare Sandboxes.
Codex now operates your computer alongside you, works with more than 90 additional plugins, generates images with gpt-image-1.5, and serves more than 3 million developers who use it every week.
Gemini 3.1 Flash TTS achieved an Elo score of 1,211 on the Artificial Analysis TTS leaderboard and supports 70+ languages with audio tags for controlling vocal style, pace, and delivery.
Sentence Transformers v5.4 adds multimodal embedding and reranker models that encode and compare text, images, audio, and video through the same API, with Qwen3-VL-Embedding-2B and Qwen3-VL-Reranker-2B supported.
OpenAI Academy published a guide on using ChatGPT for sales teams, covering account research, meeting prep, follow-ups, deal management, and measuring impact.
OpenAI is revoking and rotating its macOS code signing certificate after the Axios developer tool compromise, requiring users to update macOS apps by May 8, 2026.
Falcon Perception is a 0.6B-parameter early-fusion Transformer that reaches 68.0 Macro-F1 on SA-Co versus 62.3 for SAM 3, and Falcon OCR is a 0.3B model scoring 80.3 on olmOCR and 88.6 on OmniDocBench.
Granite 4.0 3B Vision is available on Hugging Face under Apache 2.0 and leads on PubTablesV2 cropped at 92.1 and full-page at 79.3, OmniDocBench at 64.0, and TableVQA at 88.1.
TRL v1.0 is released with a stable core following semantic versioning and an experimental layer, and the library is downloaded 3 million times a month.
OpenMed trained 4 production mRNA language models across 25 species in 55 GPU-hours for $165, with CodonRoBERTa-large-v2 reaching perplexity 4.10 and Spearman CAI correlation 0.40.
Gradio released gradio.Server, letting developers pair any custom frontend with Gradio's backend, including queuing and ZeroGPU support.
Vercel now runs lint and typecheck on every deployment in parallel with the build, available to every team, with optional required checks.
Notion rollups now support number formatting (USD, EUR, percent, etc.) and decimal places, eliminating formula workarounds.
OpenAI is launching Codex Labs and partnering with Accenture, Capgemini, CGI, Cognizant, Infosys, PwC, and Tata Consultancy Services to bring Codex to enterprises, with weekly Codex usage growing from more than 3 million to more than 4 million developers in two weeks.
Hugging Face published a guide on using Transformers.js in a Chrome extension, with a demo powered by Gemma 4 E2B for local AI features.
Hyatt has deployed ChatGPT Enterprise across its global corporate and hotel workforce, with departments including finance, marketing, and operations using it to improve guest experience.
Elastic build machines are now generally available for all Pro and Enterprise customers, and approximately 80% of projects reduced their costs by switching to smaller build machines while maintaining build speeds.
Custom Agents are now 35-50% cheaper to run, AI Autofill brings them into databases, and initial page render is 28% faster.
Claude Opus 4.7 is now in Notion, using fewer tokens and making 3x fewer tool errors than before.
Bytedance's Seedance 2.0 video generation model is available on AI Gateway in Standard and Fast variants with no provider markup.
Vercel is reducing the price of Turbo build machines by 16%, with all builds now priced at $0.0035 per CPU per minute.
Notion added a dedicated tab in settings to connect your calendar and inbox, letting Notion AI handle work across apps like scheduling and drafting emails.
Vercel Flags is now generally available, offering feature flags with targeting rules and environment controls, plus an SDK for Next.js and SvelteKit.
Notion Agent now shows your calendar to help find available time with teammates, compare schedules, and book without leaving the chat.
A high-severity React Server Components vulnerability with CVSS 7.5 can cause denial of service, and Vercel deployed WAF rules to protect all hosted projects while requiring upgrades to patched versions 15.5.15 and 16.2.3.
Claude Opus 4.6 Fast Mode is available on AI Gateway with 2.5x faster output token speeds at 6x standard Opus rates, $30 per 1M input tokens and $150 per 1M output tokens.
Vercel Sandbox now supports up to 32 vCPUs and 64 GB of RAM for Enterprise customers running large CPU-bound or memory-heavy applications.
Team-wide Zero Data Retention costs $0.10 per 1,000 requests for Pro and Enterprise teams, while per-request ZDR is free and Disallow Prompt Training is free for all users.
Stability AI launched Brand Studio, a creative production platform with Brand Central for custom Brand ID models, Producer Mode for step-by-step plans, and Precision Inpainting for targeted edits.
Notion now lets users share AI chats as read-only links, managed in Settings under Public pages and Shared AI chats.
Voice input for Notion AI is now available on desktop, allowing users to dictate prompts instead of typing.
New Vercel projects will honor cache-control headers by default when proxying requests to external origins starting April 6th, and existing projects can opt in from the project dashboard.
Vercel eliminated the $10 base fee for Observability Plus for all customers and now charges only for observability events collected.
Vercel Queues now supports a maximum message TTL and delivery delay of 7 days, up from 24 hours, while the default 24-hour TTL and 0-second delivery delay remain unchanged.
The axios npm package was compromised in a supply chain attack discovered on March 31, 2026, affecting axios@1.14.1 and axios@0.30.4, and the safe release is axios@1.14.0.
Vercel added Qwen 3.6 Plus to AI Gateway, with a 1M context window and improved agentic coding and multimodal capabilities.
Waymo opened to everyone in Miami and Orlando after welcoming over 150,000 riders from the interest list, and is introducing highway travel in Miami.
Waymo and Waze launched a pilot that uses Waymo perception systems to detect potholes and shares the data through Waze for Cities, starting in five metro areas where Waymo has identified approximately 500 potholes.
Google introduced a new approach for editing images, now live in the Auto frame feature in Google Photos, allowing users to re-imagine photos from a new perspective after they have been taken.
Google introduced ReasoningBank, a novel agent memory framework that uses successful and failed experiences to distill generalizable reasoning strategies, enabling an agent to continuously learn from experience after deployment.
Training PATHFINDER with 10% simulated data from MoGen reduced reconstruction error on reserved mouse axons by 4.4%, equivalent to saving a single expert 157 years of manual work at the scale of a complete mouse brain.
Simula generated datasets of up to 512K data points per domain, and high complexity yielded a 10% accuracy gain in math reasoning on GSM8k while hurting performance in legal reasoning on LEXam.
The edition every row came from arrives in your inbox, free.
Join free