# Mistral Moderation API vs ShieldGemma (slim)

> Mistral Moderation scores 60.6 (C) to ShieldGemma's 54 (C) for content moderation. Prices, MCP, x402, uptime and agent notes side by side.

- Full: https://www.anchorterminal.com/compare/mistral-moderation-vs-shieldgemma.md (~5,200 tokens) · this version ~2,980 tokens · JSON https://www.anchorterminal.com/compare/mistral-moderation-vs-shieldgemma.json · canonical https://www.anchorterminal.com/compare/mistral-moderation-vs-shieldgemma
- Index: https://www.anchorterminal.com/llms.txt · API: https://www.anchorterminal.com/api/v1/index.json · MCP: https://www.anchorterminal.com/mcp · Updated: 2026-11-07
- Licence: CC BY 4.0, credit Anchor Terminal and link this page (https://www.anchorterminal.com/terms/#data)

Mistral Moderation API scores 60.6 (C) on agent readiness against ShieldGemma's 54 (C), and leads in 5 of 7 scored categories. ShieldGemma leads on payments & pricing. Both do content moderation.

- Mistral Moderation API: C 60.6, rank #644 of 1486 · https://www.anchorterminal.com/tools/mistral-moderation.min.md
- ShieldGemma: C 54, rank #921 of 1486 · https://www.anchorterminal.com/tools/shieldgemma.min.md
- Mistral Moderation API, good for A free classifier for an agent that also needs PII, jailbreak and advice categories, or for a Mistral-hosted agent that can set the guardrail inline. Ahead on Schema & documentation, 88 against 56; Security & auth, 70 against 53; Maintenance & community, 83 against 40; Transparency & trust, 80 against 54. Also A hosted endpoint, with nothing to install.
- ShieldGemma, good for Policy-specific content moderation of English text input or output, or of images, on the owner's hardware, with a probability to threshold. Ahead on Payments & pricing, 52 against 40. Also No key needed to call it; No incidents deducted, where Mistral Moderation API loses 6 points for them.

| Category | Mistral Moderation API | ShieldGemma |
| --- | --- | --- |
| Reliability (16%) | 43 | 46 |
| Performance (10%) | pending | pending |
| Schema & documentation (13%) | 88 | 56 |
| Agent ergonomics (13%) | 75 | 72 |
| Security & auth (14%) | 70 | 53 |
| Payments & pricing (10%) | 40 | 52 |
| Task success (10%) | pending | pending |
| Maintenance & community (7%) | 83 | 40 |
| Transparency & trust (7%) | 80 | 54 |

| Fact (where they differ) | Mistral Moderation API | ShieldGemma |
| --- | --- | --- |
| Kind | HTTP API | Model API |
| Vendor | Mistral AI | Google DeepMind |
| Hosted endpoint | `https://api.mistral.ai/v1/moderations` | no (local only) |
| Transports | HTTP |  |
| Auth | API key | None |
| Licence | none | Gemma Terms of Use (last modified 1 April 2026) with the Gemma Prohibited Use Policy. Not an OSI licence |
| Last release | 2026-03-01 | 2025-04-04 |
| Terms last updated | 2026-09-25 | 2026-04-01 |
| Privacy policy last updated | 2026-09-03 | no document linked |
| Customer content may train models | yes, with an opt-out | yes |
| Terms restrict benchmarking | yes | not found in the text |
| Terms or service can change without notice | yes | not found in the text |
| Popularity | 769 stars, 8.4M npm/wk, 3.7M PyPI/wk | none |
| Agent reviews | 3/5 (2) | none |

Why the edge:
- Maintenance & community, Mistral Moderation API +43: Mistral Moderation API 83 (Mistral's SDKs cover every Mistral API, so their releases count here only where they change moderation code, as the Vonage and Groq listings read umbrella SDKs. Python 3.3.0 and TypeScript 3.1.0 (20 October 2026, per each repository's RELEASES.md) change the `type` field in `classifiers.moderate_chat()`, 15 days before this check (30). In the 90 days from 6 August, five SDK releases change the moderation calls, Python 3.0.0 (28 September, `mistral_mcp` across the classifier operations), Python 3.1.0 (6 October, `resource` variants in `moderate_chat()`), TypeScript 3.0.0 (7 October, `moderateChat()` inputs) and Python 3.3.0 and TypeScript 3.1.0 (20 October), each counted in full since none is packaging-only, dependency-only or documentation-only, while the changelog has no moderation entry (20). The moderation model itself last changed when `mistral-moderation-2411` retired on 30 June 2026, after deprecation on 31 March. Dated changelog and docs and support through the console (10 of 15). Current official SDKs, `mistralai` 3.7.0 and `@mistralai/mistralai` 3.5.0 (1 November) (15). SDKs generated from the OpenAPI document, with test and example workflows, as the Mistral API listing reads them (8 of 10). 30+20+10+15+8 = 83.); 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.)
- Schema & documentation, Mistral Moderation API +32: Mistral Moderation API 88 (OpenAPI document at docs.mistral.ai/openapi.yaml, per the 1 November reading on the Mistral API listing (25). llms.txt and Markdown twins of every page, re-read on 4 November (10). The guide lists the 11 categories, recommends custom guardrails over the raw API, says to use the raw score or set your own threshold and warns that model updates can need recalibration, but doesn't say when the classifier is the wrong tool or which languages it covers (13 of 20). model and input are typed, input takes a string or an array, and the chat endpoint takes typed messages (12 of 15). Python, TypeScript and curl examples, and blocked guardrail calls return 403 with the violated categories and scores (13 of 15). Dated model ids, and the changelog's 12 March 2026 entry records the release of `mistral-moderation-2603` with guardrails on chat completions, conversations and agents (15). 25+10+13+12+13+15 = 88.); 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.)
- Transparency & trust, Mistral Moderation API +26: Mistral Moderation API 80 (Closed service under commercial terms with a French legal entity, SDKs Apache-2.0 (15). Abuse logs are kept 30 days unless zero retention is granted, and `/v1/moderations` is on the zero retention list for paid plans. The terms, the DPA and the retention page agree, but the terms leave the paid API's training default unnamed and free Experiment data may be used for training, which matters because moderation is free (18 of 30, the reading the Mistral API listing gives the same pages). Model lifecycle page with notice periods per stage, and `mistral-moderation-2411` was deprecated on 31 March and retired on 30 June 2026 with both dates on its card (20). EU hosting by default and opt-in regional endpoints, with the subprocessor list in a trust centre drawn by script, read as the other Mistral listings (15 of 20). 15+18+20+15 = 68.); 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.)
- Security & auth, Mistral Moderation API +17: Mistral Moderation API 70 (Read on the same pages as the other Mistral listings. Keys are bound to one workspace, can carry an expiry date and a connector access scope, and admins can set a maximum key lifetime, re-read on 4 November. A key can't be limited to moderation (26 of 30). The same key reaches files, fine-tuning, agents, batch jobs and paid models, so there's no moderation-only or read-only key (10 of 20). A jailbreaking category in the classifier, one score rather than a document-aware injection check (12 of 15). Audit logs record API key actions on Enterprise plans only, per the 2 October reading on the Mistral API listing, and usage is shown per workspace (10 of 15). security.txt valid until 5 May 2027 with a HackerOne policy link, per the 4 November reading on the OCR listing, and the docs publish three dated security advisories for 2026, re-read on 4 November. No certification or bounty confirmed, as the trust centre is drawn by script (12 of 20). 26+10+12+10+12 = 70.); 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.)
- Payments & pricing, ShieldGemma +12: Mistral Moderation API 40 (No x402, MPP or L402 (0). Mistral Moderation 2 is listed as free ($0 per million tokens) on the pricing page, which moved to docs.mistral.ai/inference/pricing, re-read on 4 November (20). The free Experiment plan needs no card, though it needs a phone number, per the 1 October check (20). A person signs up in a browser and verifies a phone (0). 0+20+20+0 = 40.); 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.)