Claude Fable 5.2, which Anthropic calls its most capable generally available model for technical work, is now available through Serverless Inference, Inference Router (Public Preview - no SLA; may change), model synthesis, and Evaluations to eligible customers in supported regions. Anthropic reports that, in its own testing as of November 2026, it generates output up to 30% faster than Claude Fable 5.1, at $8 / $40 per million input / output tokens, lower than Fable 5.1’s $10 / $50 (USD, excluding taxes subject to change), actual results vary by workload and configuration. It runs on the same DigitalOcean account, API, and invoice you already use for serverless inference billed by DigitalOcean.

Anthropic and Claude are trademarks of Anthropic, PBC. DigitalOcean is not affiliated with, and this announcement is not endorsed by, Anthropic.

From your iPhone, Cuttlefish lets you follow live agent sessions, approve or deny tool calls with the full command or diff in view, steer or stop a run mid-flight, and review what your agent built. It works with whichever agent you run on Managed Agents, including Claude Code, Codex CLI, OpenCode, OpenHands, Hermes, LangGraph, and CrewAI.

Moonshot AI's Kimi K3.1 Pro is now available on DigitalOcean Inference Engine, a post-trained upgrade to Kimi K3 that Moonshot reports scores 65.7% on Terminal-Bench 4.0 and cuts the claim-level hallucination rate on its internal Kimi Hallu Bench Hard stress test from 23.4% (Kimi K3) to 8.3%; that benchmark is a stress test rather than an everyday error rate, and results vary by configuration and workload. It has a 1M-token context window and runs on Serverless Inference with no separate Moonshot AI account.

Moonshot AI, Kimi and Terminal-Bench are trademarks of their respective owners; no affiliation or endorsement is implied.

Team single sign-on (SSO) now supports Microsoft Entra ID, alongside Okta. If your organization runs workforce identity in Entra, your team signs in with the credentials they already use, and access follows the directory your IT team already manages.

Create, store, and manage application secrets — including API keys, database credentials, OAuth tokens, and certificates — directly from the Cloud Console, API, or CLI. Access is governed by DigitalOcean IAM, and secret activity is logged for audit without exposing secret values.

GPT-6.1 Terra, OpenAI’s fast, efficient model for visual reasoning, computer use, and natural writing, is now available through Serverless Inference, Inference Router (Public Preview), and model synthesis. OpenAI reports it performs on par with GPT-6 Astra on Agents’ Last Exam at one-tenth the price per token. It runs on the same DigitalOcean account, API, and invoice as the rest of the GPT-6 family.

See exactly what’s driving your cloud and AI spend, and catch a cost spike before it hits your invoice. Billing Insights now breaks down daily spend by model, token type, and model access key, alongside the tag, project, and product filters you already use, with CSV export and API access. A live Serverless Inference dashboard, refreshing in under 10 minutes with a 30-day lookback, is generally available today. Under the Spend Alerts tab, spend thresholds with an optional automatic pause per model access key are now in public preview.

Prompt Caching cuts the cost of repeated input — system prompts, tool definitions, conversation history — by up to 90%, with time-to-first-token up to 80% faster and no changes to your request code. Cache affinity now routes repeated requests to the replica already holding your context, lifting hit rates on flagship models like Kimi K3 and GLM-5.3 to 85–95%+. A new parameter lets you isolate cached context across teams, tenants, or end users, and a new dashboard shows your hit rate, cached vs. uncached tokens, and dollars saved, per model and over time.

Simulations, a new capability of DigitalOcean Evaluations, lets you test a change through realistic multi-turn scenarios in an isolated sandbox on Harness Runtime, with the same configuration and Action Gateway tools your agent uses in production. Pull scenarios from Signals, generate them from your agent’s configuration, or describe one in plain language. An LLM judge scores every trajectory, and running baseline against candidate shows exactly what improved and what regressed.

Run a Simulation ->

Signals watches production traffic on Managed Agents and Inference Engine and automatically highlights potential problems across user sentiment, tool execution, and infrastructure, with nothing extra to instrument. Flagged sessions become evaluation scenarios in one click. Signals is opt-in, designed to redact PII before processing, and keeps traces within your DigitalOcean Team boundary.

Enable Signals ->