Azure Translator vs Localazy

Azure Translator 75

Model reading of the checklist, with training and retention in place of the least-privilege and injection lines. Resource keys, 10-minute bearer tokens, or Microsoft Entra ID with the Cognitive Services User role and managed identities (30), less 10 because the authentication reference documents passing the key in a Subscription-Key query parameter (20). The data privacy page says text translation doesn't store customer data, which leaves nothing to train on, but we didn't find an explicit training statement there (15 of 20). Text isn't stored and documents are hard-deleted after processing. LLM translation runs through your own Foundry deployment, whose retention terms we didn't check (12 of 15). Azure Monitor and the activity log record resource actions, per-request logging not confirmed (10 of 15). MSRC disclosure policy, Azure bounty programme, SOC 2 and ISO 27001 reports. The microsoft.com security.txt passed its Expires date on 2026-09-23 and still carries it on 1 November. An MSRC search on 1 November lists no CVE for the Translator service (18 of 20).

Localazy 37

Bearer tokens in the Authorization header. A project token reaches every endpoint of one project with read and write access; a translation token reaches only the AI Translation endpoint; organisation tokens are in beta on request (20). The translation token is the only least-privilege option, and project tokens can delete keys and glossary items without a confirmation step (10). The API returns strings written by translators and contributors, with no guidance on untrusted content (5). No per-call log or audit trail for tokens found (2). localazy.com/.well-known/security.txt and /security return 404, and we found no disclosure policy or certification (0). 20+10+5+2+0 = 37.

Azure Translator 83

OpenAPI documents for text translation 3.0 and 2026-06-06 are public in Azure/azure-rest-api-specs, with examples on each operation (25). No llms.txt found on learn.microsoft.com (0). Operation descriptions state purpose, and the overview says to pick NMT or an LLM by quality, cost and scenario, but not when to avoid either (15 of 20). Enums for profanityAction, textType, tone and gender, required fields marked, but the LLM deploymentName is a free string (13 of 15). Examples in the spec and a status-code page with six-digit error codes and their meanings (15). Dated api-version values and a What's New page (15).

Localazy 51

No OpenAPI or other machine-readable contract was found (0). localazy.com/llms.txt is a six-link index of top-level pages, with no Markdown docs for agents (5). Each endpoint page states which token types can call it, whether write permission is needed and the minimum role, and the limits page says to use the CDN, not the API, for production fetches (14). Request bodies list required and optional fields with limits, such as 512 characters for a key and 1,000 for a comment (12). Sample requests, responses and error responses per endpoint (13). Docs pages carry 'last updated' dates and the blog posts dated release updates, but there's no API version or changelog (7). 0+5+14+12+13+7 = 51.

Azure Translator 78

Closed service under Microsoft's product terms, SDKs under MIT (15). The data privacy page and privacy statement agree on no storage of text and hard deletion of documents, but say nothing on training, and the LLM path falls under separate Foundry terms (22 of 30). Microsoft's lifecycle policy applies, but the What's New page gives no retirement date for v3.0 even though 2026-06-06 breaks its request shape (10 of 20). Regional resources and endpoints for data residency and a public Microsoft sub-processor list (18 of 20).

Localazy 57

Editorial half only. Closed service under terms that name Localazy s.r.o. of Brno and Czech law; the CLI is BSD-3-Clause and the API client MIT (15). The privacy policy keeps data while the service is used and 24 months after, and includes a DPA, but says nothing on whether content sent to Localazy AI is used for training (15). No deprecation policy or dated notices found (0). The DPA links a list of suppliers and subprocessors on an external knowledge base we didn't read (10). 15+15+0+10 = 40.

Azure Translator 82

API reading of the checklist. Responses hold one translation per target with optional alignment and sentence lengths, and up to 1,000 texts go in one NMT call, but there's no field selection (20 of 25). Several targets per call, a scope filter on the language list, and top and skip paging on document jobs (18 of 20). Six-digit error codes such as 400036 (invalid target language) and 403001 (free quota exceeded) say what to fix (20). Text translation is stateless, but there's no 429 retry guidance and no idempotency key on document batches (12 of 20). A key, a region header and api-version on every call, with SDKs in Python, JavaScript, .NET and Java (12 of 15).

Localazy 62

API reading. List endpoints return whole projects, files and keys with few output controls (13). Filtering by file, language and tags, and batch imports of up to 1,000 keys and 50 files a request (14). Error responses are documented per endpoint, with HTTP codes and messages (12). No idempotency keys, and the docs warn against parallel imports into the same new file (8). Official TypeScript API client, a CLI and Android and iOS libraries, with few required fields (15). 13+14+12+8+15 = 62.

Azure Translator 83

Azure status page with post-incident reviews (20). The 90-day window now starts on 3 August. No review in it names Translator. The 29 September 2026 review covers intermittent failures and 500 and 503 errors for Azure OpenAI, Foundry and Azure AI services in Sweden Central from 10:03 to 15:58 UTC, which may touch Translator resources and LLM deployments there, one region and partial, so minor. The 30 September to 1 October gateway review covers networking services and doesn't name Translator. The public page lists only broad incidents (20). Limits published with numbers, 1,000 texts and 50,000 characters a request and hourly caps from 2 million characters on F0 to 200 million on S4 (15). 429 has three documented sub-codes (429000 to 429002) and the limits page asks for load spread across the hour, but there's no Retry-After or backoff guidance for 429, re-read on 1 November. Retry advice exists only for 408 and 503. Text translation has no side effects, so a retry is safe (8 of 15). The old Cognitive Services SLA address now redirects to Microsoft's Online Services SLA, a download we didn't open (10). API 2026-06-06 is GA (10). Same window reading as the Azure Speech, Key Vault and Blob Storage listings on 1 November.

Localazy 68

Hosted lines, for the REST API at api.localazy.com. Better Stack status page with Web, API, Public API, Payments and Connect API components and 30, 60 and 90-day uptime (20). The incidents page reports none for September, October or November 2026, but the Public API shows 99.805% uptime in one 30-day bucket, roughly an hour and a half of downtime with no incident posted, so the line is held below clean (25). Limits per plan of 100 requests a minute and 10 a second, plus daily import and fetch caps per project (15). No 429 or retry guidance found in the API docs (0). No SLA found on the pricing page or in the terms (0). The core API is GA; the AI Translation endpoint is marked beta (8). 20+25+15+0+0+8 = 68.

Azure Translator 20

No x402, MPP or L402 (0). Per-million-character prices published without a login, readable through the Azure Retail Prices API since the pricing page needs JavaScript (20). F0 gives 2 million characters a month, but an Azure subscription needs a card (0). A person signs up in a browser (0).

Localazy 35

No x402, MPP or L402 (0). Plan prices are public ($34, $78 and $175 a month billed annually) and credits are $1 each, with Localazy AI at about $0.005 a word on average, so partial per-unit credit (15). A free plan for small projects and a 14-day Business trial with no credit card (20). A person signs up in a browser and mints tokens in the developer settings (0). 0+15+20+0 = 35.