Google Cloud Model Armor vs ShieldGemma

Google Cloud Model Armor 85

Newest release note 29 October 2026, Model Armor in southamerica-east1 (São Paulo) (30). 14 dated release notes between 3 August and 29 October 2026, 23 since 8 June (20). Public release notes and Cloud Customer Care. The client libraries link a Google Issue Tracker component for Model Armor, which we couldn't read (12 of 15). Official client libraries for Python and Node.js on PyPI and npm (15). @google-cloud/modelarmor 0.11.0 was tagged on 21 October 2026 and google-cloud-modelarmor 0.7.4 on 29 October 2026, both from monorepos with CI, but both are 0.x and marked release_level preview, and npm and PyPI didn't resolve from our shell on 1 November (8 of 10).

ShieldGemma 40

Read as a model, on the lines used for Granite Guardian (recency 30, deprecation 12, churn 8, responsiveness 25, packages 15, health 10). The newest file change is ShieldGemma 2 on 4 April 2025, 19 months before the check, and ShieldGemma 1 last changed in September 2024 (0 of 30). No deprecation policy. Both generations stay downloadable, and the ai.google.dev page was marked deprecated with a pointer to gemma.ai without dates for the models (3 of 12). Two generations in 2024 and 2025, each in its own repository, none replaced in place (6 of 8). Google accounts (members of Google's Gemma organisations on the Hub) answered every thread we sampled on the 2B, 9B and ShieldGemma 2 pages, from August 2024 to 22 March 2026, but leave most threads open (16 of 25). Transformers, Keras and Ollama's library load it (12 of 15). No CI, and Ollama's library build was last updated 2 years ago (3 of 10). 0 + 3 + 6 + 16 + 12 + 3 = 40.

Google Cloud Model Armor 90

Model Armor is its own product on status.cloud.google.com (20). Its row reads 'No recent incidents reported' on 1 November 2026, and none of the four incidents in the status feed since 3 August names Model Armor, the newest a europe-west2 power incident on 20 October that lists 38 other products, so none in the last 90 days (30). 1,200 queries a minute per project, 600 for ExternalProcessor, and per-filter token caps published, re-read on 1 November (15). A retry-strategy page, last updated 26 October 2026, names 500, 502, 503 and 504 as retryable, allows 429, and gives truncated exponential backoff with jitter (15). Model Armor isn't on the Google Cloud SLA list, re-read on 1 November (0). The two screening methods are GA, image screening is preview (10).

ShieldGemma 46

Scored on the local-software lines, since the owner runs the model, the reading used for Llama Guard and Granite Guardian. Google's install line is pip install -U transformers[accelerate] with no versions, a Keras Colab notebook covers ShieldGemma 1, and Ollama's own library carries ShieldGemma 1. No Python version or memory figure is stated, and the publisher's files sit behind a login and licence click-through (12 of 20). No CI, tests or evaluation code for the models were found. The technical reports describe the evaluation and nothing reruns it (0 of 25). On the Hub, 8 of 23 threads on the 2B, 2 of 10 on the 9B, none of 8 on the 27B and 9 of 10 on ShieldGemma 2 are open, but every open thread we sampled has a reply from a Google account, the newest on 22 March 2026 (15 of 25). No changelog or version tags. The two generations are named and sit in their own repositories (4 of 15). Released with no preview label (15 of 15). 12 + 0 + 15 + 4 + 15 = 46.

Google Cloud Model Armor 90

OAuth 2.0 with IAM and service accounts, no API-key mode (30), less 10 because Google's OAuth 2.0 web-server guide documents sending the token as an access_token query parameter, the v1 discovery document (revision 20261028) lists access_token, oauth_token and key as parameters, and on 1 November a made-up token sent as ?access_token= to modelarmor.googleapis.com and modelarmor.us-central1.rep.googleapis.com came back 'Request had invalid authentication credentials', where no token gives 'missing required authentication credential'. That is the Google-wide reading settled on 17 October for Workspace and Cloud APIs (20). Each screening method has its own permission, such as modelarmor.templates.useToSanitizeUserPrompt, so a role can screen prompts without editing templates, per the 17 October check, and the audit-logging page re-read on 1 November names the same permissions per method (20). Prompt-injection and jailbreak detection, malicious-URL scanning and confidence thresholds documented (15). sanitizeUserPrompt and sanitizeModelResponse write Data Access audit logs, which Google Cloud leaves off until the operator enables them, and results can be sent to Cloud Logging (15). google.com security.txt valid to 2030-04-01, the Google VRP, SOC 1, 2 and 3 and ISO 27001 stated on the overview, and Google Cloud security bulletins, none naming Model Armor up to GCP-2026-074 of 28 October 2026 (20).

ShieldGemma 53

Read as a model the owner runs, on the model lines used for Granite Guardian, with retention and training in place of the read-only line. No key is needed to run it. Google's files need a Hugging Face login and licence acceptance, approved immediately, so there is a token to download but no secret at run time (13 of 30). Self-hosted, so inputs stay on the owner's hardware and nothing reaches Google, though no Google page says so (15 of 20). It reads untrusted text by design. The card says the model is highly sensitive to the policy wording and may behave unpredictably with ambiguous language, and gives no prompt-injection guidance (6 of 15). No call log of its own, though each call returns a score (5 of 15). Weights are safetensors only and load through Transformers classes with no remote code. Google's security.txt is valid and points to its reward programme, and there is no model-specific security page or advisory record (14 of 20). 13 + 15 + 6 + 5 + 14 = 53.

Google Cloud Model Armor 87

Closed service under the Google Cloud terms, which set the contracting Google entity by customer region (15). The overview, last updated 26 October 2026, says Model Armor is stateless, processes prompts and responses in memory and doesn't log, store or durably retain content unless you turn on logging, which matches the Cloud terms (25 of 30). Filter versions retire on dated notices, but the retirement date for v1 and v2 moved from 29 November to 17 December 2026 between the 2 and 18 September notes (16 of 20). Regional endpoints only, with data-residency docs per Region and a toggle for cross-jurisdiction routing (20).

ShieldGemma 54

The weights are under the Gemma Terms of Use, not an OSI licence. They carry distribution conditions, the Prohibited Use Policy and a clause that lets Google restrict usage remotely where it believes the terms are breached. The terms are dated (1 April 2026) with earlier versions archived (15 of 30). Inputs stay on the owner's hardware, so no privacy statement applies. ShieldGemma 1 was fine-tuned on synthetic internal data and public datasets described in a technical report, and ShieldGemma 2 on a WebLI subset and synthetic images with CSAM filtering. No data is released (20 of 30). No deprecation policy or dated model retirements (5 of 20). The weights carry no telemetry of their own, with no statement either way, and downloads go through Hugging Face, Kaggle or Ollama (10 of 20). 15 + 20 + 5 + 10 = 50.

Google Cloud Model Armor 20

No x402, MPP or L402 (0). $0.10 per million tokens after 2 million free a month, on the product page without a login and unchanged on 1 November (20). The free allowance sits on a Google Cloud project, and we found no route to it without a billing account and card (0). A person creates the project, enables the API and sets up IAM in a browser (0).

ShieldGemma 52

Self-hosted rule. Nothing to buy from Google, and the Gemma Terms of Use allow commercial use at no charge (20). Free to run with no card (20). Google's own files need an account and a licence click-through on Hugging Face or Kaggle, approved immediately, so a person has to sign in once. Ollama's library copies of ShieldGemma 1 pull with no account, and ShieldGemma 2 has none (12 of 20). No payment protocol (0). No hosted ShieldGemma from Google or an inference provider was found. 20 + 20 + 12 + 0 = 52.

Google Cloud Model Armor 78

Public discovery document for modelarmor v1, revision 20261028 on 1 November, with typed parameters, resource-name patterns and enums (25). No llms.txt, and docs.cloud.google.com/llms.txt returned 404 on 1 November (0). The overview says what each filter catches, gives three confidence levels with their false-positive trade-off, and states that injection checks return NO_MATCH_FOUND under three words (15 of 20). Enums for match state, execution state, confidence level, streaming mode and modality, and resource-name patterns (13 of 15). Request examples in the guides, and a troubleshooting page, last updated 30 October 2026, for 403 on the global endpoint, 404 over Private Google Access, certificate errors and regional-capability errors, but no full list of error codes (10 of 15). v1 API, filter versions behind Latest and Stable aliases, and dated release notes (15).

ShieldGemma 56

Read for a model the owner serves. There is no specification of its own. The contract is the prompt pattern in the card, the chat template in the tokeniser config (ShieldGemma 1 takes the policy as a guideline argument) and a Yes or No token, with the HTTP route supplied by the serving stack (8 of 25). gemma.ai/llms.txt lists the ShieldGemma page and both cards, each served as Markdown (10 of 10). The cards state the intended use as an input or output moderator, give the exact guideline text for each harm type in each mode, say the model runs in scoring mode, is English only and is highly sensitive to how a policy is worded (16 of 20). Fixed harm types and a binary answer, with the policy as free text (10 of 15). A Transformers snippet for each generation and a Keras notebook. No failure modes or error cases are documented (8 of 15). Two named generations with no changelog, and the ai.google.dev page is marked deprecated in favour of gemma.ai (4 of 15). 8 + 10 + 16 + 10 + 8 + 4 = 56.