- Contentdigitalocean.com
Request Queuing and Latency Under Burst Traffic: What Happens When Your Model Gets Overloaded?
DigitalOcean released a technical tutorial explaining how to measure and manage request queuing and latency during burst traffic for LLM inference workloads. The guide covers testing on Serverless Inf…
- Contentdigitalocean.com
How do I monitor and trace AI agents running in Production
DigitalOcean demonstrates how traditional logging fails to catch silent failures in AI agents, where tool calls succeed but produce wrong answers. Using DigitalOcean Managed Agents, the company shows …
- ideas.digitalocean.com
Now Available: Billing Insights
DigitalOcean introduced Billing Insights, a tool that provides daily spend breakdowns by model, token type, and access key, with tag, project, and product filters, CSV export, and API access. It inclu…
- digitalocean.com
Inside DigitalOcean Simulations: how user simulation, execution, and validation work
DigitalOcean introduced Simulations, a Public Preview feature in its Evaluations product, to test AI agents through multi-turn conversations rather than single prompts. Simulations records full trajec…
- Featuredigitalocean.com
How to Run Untrusted LLM-Generated Code in an Isolated Environment With MicroVMs
DigitalOcean released MicroVMs in public preview on October 29, 2026, as a secure sandbox for running untrusted LLM-generated code. Built on Firecracker, each MicroVM runs in its own isolated VM with …
- digitalocean.com
Grok Bot vs Dots vs Muse: Which Agent Fits Developers in 2027?
DigitalOcean published a detailed comparison of Grok Bot (xAI), Dots (OpenAI), and Muse (Meta) as always-on AI agents, emphasizing their lack of documented APIs for programmatic control. The article p…
- digitalocean.com
DigitalOcean Signals: Observability for AI Agents and Inference
DigitalOcean launched Signals, a Public Preview feature that analyzes production AI agent conversations to detect behavioral signals like user frustration, looping, tool failures, or rate limits. Sign…
- Featuredigitalocean.com
DigitalOcean Managed Agents latency: cold start and resume
DigitalOcean’s Managed Agents resume latency was initially misreported as 540 ms due to polling artifacts, but corrected measurements show 249 ms at the median. New sessions take ~650 ms to become rea…
- docs.digitalocean.com
28 October 2026 (managed-agents, agent-harness-runtime)
DigitalOcean released Cuttlefish, an iOS app for its Harness Runtime, now in public preview on the App Store. Users can follow and steer sessions, approve or deny tool calls, and manage workspace file…
- Launchapps.apple.com
Cuttlefish by DigitalOcean
DigitalOcean released Cuttlefish, an iOS app enabling users to run and manage AI coding agents (Claude Code, Codex, OpenCode, OpenHands, etc.) directly from iPhones. The app supports team collaboratio…

- digitalocean.com
9 Best Langfuse Alternatives for LLM Observability in 2027
DigitalOcean introduced DigitalOcean Insights, a built-in observability tool for its AI products (Serverless Inference, Batch Inference, Managed Agents, etc.) that requires no SDK or collector setup. …
- Featuredigitalocean.com
Agents in Production, Part 3: Let Your AI Support Agent Send Replies, Without Giving It the Credential
DigitalOcean’s AI support agent now uses Action Gateway to broker credentials, removing API keys from the agent’s environment and preventing credential leaks. The system triaged 335 tickets, with 57 r…
- Featuredigitalocean.com
Three layers, all config: how DigitalOcean Managed Agents enforce NVIDIA OpenShell-aligned boundaries
DigitalOcean introduced Managed Agents in public preview, enabling AI agents to perform production tasks like migrating a backend from Supabase to DigitalOcean without direct credential access. The pl…
- digitalocean.com
What Is Batch Inference? (2027)
DigitalOcean introduced batch inference on its Inference Engine, allowing users to process large AI workloads asynchronously at up to 50% lower cost than real-time rates. The feature supports OpenAI a…
- Contentdigitalocean.com
Cold Start Mechanics: How Warm Pools and Model Caching Actually Work
DigitalOcean’s tutorial breaks down why AI inference services experience delays (cold starts) when scaling, attributing it to model loading times on GPUs. It details how warm pools and model caching r…
- digitalocean.com
Whats is Batch Inference? (2027)
DigitalOcean introduced Batch Inference on its Inference Engine, enabling asynchronous processing of large AI workloads at up to 50% lower cost than real-time rates. The feature supports OpenAI and An…
- Sunsetdocs.digitalocean.com
26 October 2026 (inference)
DigitalOcean’s Inference service will stop supporting OpenAI’s o1, o3-mini, and GPT Image 1 models effective 26 October 2026. Users are directed to migrate to GPT-5.6 Sol for o1/o3-mini and GPT Image …
- Securitydocs.digitalocean.com
25 October 2026 (managed-agents, agent-harness-runtime)
DigitalOcean rebuilt Harness Runtime platform templates to address CVE-2026-111497, a remote code execution vulnerability in the Brotli decoder affecting Node.js. The update impacts templates like cod…
- Featureideas.digitalocean.com
Now Available: Serverless Inference Prompt Caching
DigitalOcean introduced serverless inference prompt caching, reducing costs for repeated inputs by up to 90% and cutting time-to-first-token by up to 80%. The feature includes a new parameter for cach…
- minimax.io
MiniMax M3.1 Flash: Redefining Agent Performance and Cost
MiniMax unveiled M3.1 Flash, a next-generation AI model optimized for agentic tasks, coding, computer use, and cybersecurity. It delivers nearly three times faster online inference and reduces usage c…
- Featuredigitalocean.com
DigitalOcean launched Signals in Public Preview, an observability tool that analyzes production AI agent conversations to detect 25 signal types across interaction, execution, and environment layers.
- digitalocean.com
Prompt Caching is GA on DigitalOcean Serverless Inference: Up to 90% Off the Context Your Agents Resend
DigitalOcean made prompt caching generally available for its Serverless Inference platform, reducing costs by up to 90% for cached input tokens and cutting time to first token by up to 80%. The featur…
- docs.digitalocean.com
23 October 2026 (inference)
DigitalOcean introduced MiniMax M3.1-Flash in public preview for its Inference serverless AI service. The model is now available alongside existing options, expanding DigitalOcean’s AI inference capab…
- Featuredocs.digitalocean.com
23 October 2026 (billing)
DigitalOcean enhanced its Cost Analysis page with new grouping and filtering options for spend tracking, including project, tag, model, token type, and model access key. These options are now availabl…
- Featuredocs.digitalocean.com
22 October 2026 (app-platform)
DigitalOcean updated its PHP buildpack for App Platform to version 298, adding support for PHP versions 8.5.3–8.5.12, 8.4.18–8.4.27, 8.3.31–8.3.33, 8.3.35–8.3.36, and 8.2.31–8.2.35. The update expands…
- Securitydigitalocean.com
Critical PHP RCE Vulnerabilities: Patch Now (CVE-2026-109618, CVE-2026-109619)
DigitalOcean disclosed two critical PHP remote code execution vulnerabilities (CVE-2026-109618 and CVE-2026-109619) on October 21, 2026, with public exploit code available. The flaws affect all PHP ve…
- Technicaldigitalocean.com
Product Security in the Post-AI Era
DigitalOcean’s Product Security team acknowledges that AI has fundamentally disrupted traditional software development lifecycles, replacing linear phases with a chaotic, agent-driven 'Jeremy Bearimy …
- nodejs.org
Wednesday, October 21, 2026 Security Releases
Node.js released security updates on October 21, 2026, fixing 11 vulnerabilities across versions 26.x, 24.x, and 22.x, including high-severity issues in DNS resolution, TLS hostname verification, Brot…
- Featuredocs.digitalocean.com
21 October 2026 (inference)
DigitalOcean introduced the DeepSeek V4.1 Flash 1017 model to its Inference service, enabling serverless AI model deployment. The model is now available for immediate use alongside existing options.…
- Partnershipdigitalocean.com
How Workato’s AI Research Lab Is Scaling the Agentic Enterprise with Optimized NVIDIA H200 Clusters from DigitalOcean
DigitalOcean partnered with Workato AI Research Lab to deploy NVIDIA H200 GPU clusters and DigitalOcean Managed Kubernetes, enabling 2-3x faster time-to-value and 67% lower inference costs. The collab…
- Case Studydigitalocean.com
LawVo Scales AI-Powered Legal Services with DigitalOcean
LawVo, a Canadian legal tech startup, chose DigitalOcean for its Toronto-based cloud infrastructure to meet data residency needs while scaling AI-powered legal services. The company leverages DigitalO…
- Integrationdigitalocean.com
ACE Studio Achieves Real-Time AI Music Creation with DigitalOcean and AMD
ACE Studio, an AI-native music workstation, migrated to DigitalOcean GPUs powered by AMD Instinct MI325X to address latency and cost challenges, reducing inference latency by 80% and cutting training …
- Case Studydigitalocean.com
From Data to Discovery: How Traversal Accelerates AI Innovation on DigitalOcean
Traversal, an AI SRE startup, chose DigitalOcean for its developer-friendly infrastructure to power root cause analysis for production systems. The company relies on DigitalOcean Kubernetes, GPU Dropl…
- Case Studydigitalocean.com
How Consumable scales AI with DigitalOcean Inference Router
Consumable, an adtech company handling up to one million queries per second, migrated from AWS to DigitalOcean in 2021 for cost predictability and scalability. They now use DigitalOcean’s Serverless I…
- Case Studydigitalocean.com
NoBid’s migration from AWS to cost-effective cloud performance
NoBid, an ad-tech startup processing 200 billion auctions monthly, migrated its infrastructure from AWS to DigitalOcean in late 2024 to reduce costs and improve performance. Using DigitalOcean’s Kuber…
- Case Studydigitalocean.com
How Probably Delivers Verifiable Data Analysis on DigitalOcean
Probably, an AI agent startup, reduced infrastructure costs by 25% and deployed its API in 1.5 days on DigitalOcean’s AI-Native Cloud, leveraging App Platform for orchestration and edge networking. Th…
- Case Studydigitalocean.com
RadixArk serves DeepSeek V4 at 3.5K+ tokens/sec/GPU on DigitalOcean and AMD
RadixArk leveraged DigitalOcean’s AMD Instinct MI350X/MI355X GPU Droplets to deploy DeepSeek V4 at over 3,500 tokens/sec/GPU, achieving ~10x throughput gains through joint AMD engineering. The setup e…
- Case Studydigitalocean.com
Scaling to Billions: Traject Data delivers lightning-fast web data with DigitalOcean VPNs
Traject Data, a SERP and e-commerce data API provider, scaled to 2 billion monthly requests using DigitalOcean's Droplets, VPCs, Spaces, and Managed Databases. The partnership enabled cost-effective o…
- Partnershipdigitalocean.com
How Hippocratic AI Holds a 99.9% Safety Score Across 10 Million Patient Calls on DigitalOcean's AI-Native Cloud
DigitalOcean collaborated with Hippocratic AI and NVIDIA to deliver an AI-Native Cloud stack that achieved a 99.9% clinical safety score across 10 million patient calls. The partnership enabled 2x pro…
- Contentdigitalocean.com
7 Replicate Alternatives for AI Inference in 2027
DigitalOcean published a guide comparing Replicate alternatives for AI inference in 2027, highlighting Replicate’s cost unpredictability and cold-start issues. The article positions DigitalOcean’s AI-…
- digitalocean.com
Building a computer that can pause and resume in milliseconds: Inside the MicroVM architecture behind Managed Agents
DigitalOcean unveiled a MicroVM-based architecture for its Managed Agents, leveraging Firecracker to enable near-instant pause/resume and checkpoint restore of agent environments. The system uses cell…
- Featuredocs.digitalocean.com
15 October 2026 (inference)
DigitalOcean introduced NVIDIA’s Nemotron 3.5 Super VL model to its Inference service, enabling serverless inference for users. The model is now listed on the Available Models page for immediate use.…
- Featuredigitalocean.com
Inside Action Gateway in Managed Agents
DigitalOcean unveiled Action Gateway, a new capability within Managed Agents that governs external tool calls made by AI agents. It enforces policy-based authorization, human approval workflows, and c…
- digitalocean.com
11 Best Open-Source LLM Hosting Providers in 2026
DigitalOcean now offers open-source LLM hosting via a serverless API supporting models like DeepSeek, Llama, and Qwen, with 1-Click Model deployments and dedicated inference hardware (AMD MI300X to NV…
- Featuredigitalocean.com
DigitalOcean Inference Router Enhancements: Performance Tuned Routing, No Rules to Write
DigitalOcean enhanced its Inference Router to automatically route prompts to the best-performing model from its catalog, eliminating manual configuration. The router uses public benchmarks and interna…
- Featureideas.digitalocean.com
Now in Public Preview: DigitalOcean Signals
DigitalOcean launched Signals in public preview, a new feature that automatically monitors production traffic on Managed Agents and Inference Engine to highlight potential issues in user sentiment, to…
- Featureideas.digitalocean.com
Now in Public Preview: Simulations
DigitalOcean introduced Simulations, a new feature in its Evaluations product, enabling multi-turn scenario testing in an isolated sandbox using Harness Runtime. Users can generate scenarios from Sign…
- ideas.digitalocean.com
Now in Public Preview: DigitalOcean Insights
DigitalOcean launched Insights in public preview, offering a unified platform to detect, investigate, and monitor metrics, logs, and traces across applications and infrastructure. The feature includes…
- digitalocean.com
Introducing DigitalOcean Insights: Built-in Observability for Your Agents, Applications, and Infrastructure
DigitalOcean introduced Insights, a new built-in observability tool that consolidates metrics, logs, and traces from DigitalOcean Kubernetes, App Platform, and agent services into a single interface. …
- Launchdigitalocean.com
Introducing DigitalOcean Agent Data Services: The Data Foundation for Agents
DigitalOcean introduced Agent Data Services, a managed data foundation for AI agents, now in private preview. The service unifies multimodal data ingestion, vector indexing, and execution traces into …
- docs.digitalocean.com
12 October 2026 (evaluations)
DigitalOcean introduced Simulation, a public preview feature that tests AI agent behavior through multi-turn conversations using scenario sets, user simulators, judge models, and conversational metric…
- Featuredigitalocean.com
Introducing DigitalOcean Signals and Simulations: see how your agents and models really behave in production, then test the fix before it ships
DigitalOcean introduced Signals and Simulations as part of its Agent Learning Services to address AI agent failures in production. Signals automatically detects issues like wrong tool usage or user fr…
- digitalocean.com
DigitalOcean MicroVMs: Fast, isolated compute for your AI agent infrastructure
DigitalOcean introduced MicroVMs in public preview, offering fast, isolated compute built on Firecracker for AI agent platforms. MicroVMs pause when idle and resume instantly, reducing costs by chargi…
- docs.digitalocean.com
9 October 2026 (microvms)
DigitalOcean introduced MicroVMs in public preview within its MKC1 region, offering lightweight VMs that run container images with their own kernels, auto-pause when idle, and resume with preserved st…
- ideas.digitalocean.com
Now Generally Available: Organization-Level Resource Limits
DigitalOcean introduced organization-level resource limits, allowing Owners and Admins to view and adjust team-level limits for Droplets, Volumes, Databases, and Load Balancers in a single interface. …
- Featureideas.digitalocean.com
Now Generally Available: Configurable Session Durations
DigitalOcean introduced configurable session durations, allowing users to set login sessions from 1 hour to 30 days based on device trust levels. Sessions persist for the full duration without early c…
- Featuredocs.digitalocean.com
8 October 2026 (kubernetes)
DigitalOcean Kubernetes (DOKS) now automatically enables the CoreDNS Cluster Proportional Autoscaler (CPA) plugin when clusters are upgraded to version 1.37 or later, provided CoreDNS is running the d…
- Featuredocs.digitalocean.com
DigitalOcean Kubernetes (DOKS) 1.37 no longer enables GPU device plugins by default on clusters with GPU nodes. Instead, Dynamic Resource Allocation (DRA) drivers for NVIDIA and AMD GPUs are available
- Featuredigitalocean.com
Account Control, Every Layer: From Your Organization to Every Session
DigitalOcean introduced two new controls: Resource Limits for organization Owners and Admins to centrally manage and adjust resource caps across all teams, and configurable session durations (1 hour t…
- Launchdocs.digitalocean.com
7 October 2026 (postgresql, mysql, mongodb, valkey, opensearch, kafka)
DigitalOcean introduced managed databases (PostgreSQL, MySQL, MongoDB, Valkey, OpenSearch, Kafka) in its Memphis (mem1) region, available exclusively on AMDv5 plans. Storage-Optimized plans are now al…
- Featuredocs.digitalocean.com
7 October 2026 (inference, billing)
DigitalOcean introduced a new Cost Analysis tab feature that tracks Serverless Inference Live Usage, displaying estimated costs and token counts by model and access key for teams and organizations. Th…
- Contentdigitalocean.com
DigitalOcean vs Baseten: AI Inference Comparison (2026)
DigitalOcean published a detailed comparison of its AI inference platform against Baseten, emphasizing its pay-per-token billing model versus Baseten’s per-replica-minute pricing. The article highligh…
- Featuredocs.digitalocean.com
6 October 2026 (managed-agents, agent-harness-runtime)
DigitalOcean’s Harness Runtime now supports deploying CrewAI agents, enabling users to package CrewAI repositories, define inputs, connect models and tools via Action Gateway, and run crews in isolate…
- reflection.ai
Beam is now available
Reflection AI released Beam, a 501B-parameter sparse Mixture-of-Experts model under Apache 2.0, optimized for coding, reasoning, and long-horizon agentic tasks. The model is available on Hugging Face …

- ideas.digitalocean.com
Introducing Agent Droplets
DigitalOcean introduced Agent Droplets, a new compute offering that consolidates compute, inference, storage, and tools for AI agents into a single plan. Pro and Team plans receive 15% and 20% discoun…
- Featuredocs.digitalocean.com
doctl serverless-inference
DigitalOcean introduced the 'doctl serverless-inference' CLI tool in version 1.177.0, enabling direct calls to its serverless inference API at inference.do-ai.run. Users can authenticate with model ac…
- digitalocean.com
Introducing Agent Droplets: everything an agent needs, one price, one bill
DigitalOcean introduced Agent Droplets, a bundled pricing model for its Managed Agents service, consolidating compute, inference, storage, and tool access into a single monthly plan. This simplifies b…
- Integrationideas.digitalocean.com
Now Available: GPT-6.1 Sol from OpenAI
DigitalOcean now offers OpenAI’s GPT-6.1 Sol model through its Serverless Inference platform, providing near-GPT-6 Astra performance for coding, computer use, and document tasks at lower per-token cos…
- Featuredigitalocean.com
DigitalOcean MicroVMs: the compute foundation for building your AI agent infrastructure (private preview)
DigitalOcean introduced MicroVMs, isolated virtual machines that pause when idle and resume with state intact, designed for AI agent workloads and other bursty, isolated compute needs. Powering their …
- Pricingdocs.digitalocean.com
1 October 2026 (billing, inference, managed-agents)
DigitalOcean launched a dedicated Inference and Agents balance for prepaid Serverless Inference and Managed Agents, separate from other product balances. Users can add $5–$500 or enable auto-reload vi…
Track DigitalOcean on autopilot
- · Weekly AI brief: narrative summary of what shipped, every Monday 9 AM
- · Email or Slack alerts, or chat with the archive in your dashboard
- · Add DigitalOcean + up to 2 more competitors free, no credit card