Amazon Bedrock vs Gemini Developer API

Amazon Bedrock 98

Model reading. IAM with SigV4, roles and STS session credentials, policies down to a model or inference-profile ARN, short-term API keys that expire within 12 hours, long-term keys with expiry and reset, and condition keys that can deny bearer tokens altogether (30). The security page says customer data isn't shared with model providers or used to improve the base models (20). The abuse detection page says Bedrock stores no inputs or outputs by default, and a per-Region none mode blocks models that require retention, but Claude Fable 5 and 5.1 keep all traffic up to 30 days and GPT-6 and GPT-5.x keep flagged traffic up to 30 days (13 of 15, a judgement call; Anthropic's own API got 12 for the same Fable exception without an account-level block). CloudTrail logs API calls, and model invocation logging can send prompts and responses to S3 or CloudWatch Logs (15). Bedrock is in ISO, SOC and CSA STAR Level 2 scope, HIPAA eligible and FedRAMP High in GovCloud (US-West) per its security page, with a vulnerability disclosure programme on HackerOne, and the aws.amazon.com security.txt, expired on 6 October, was renewed with Expires 2027-10-16 when read on 6 November (20). Up from 96 on 6 October for that reason.

Gemini Developer API 50

Model reading. Keys live in a Cloud project. Since 28 May 2026 new keys are authorisation keys bound to a service account and restricted to the Gemini API, keys can be limited by IP or origin, unrestricted standard keys are rejected, and the docs give a leak checklist, per the API key page re-read on 1 November. No permission scopes, so 25 of 30, less 10 because the API reads a token from the URL. The Discovery document (revision 20261101) lists access_token, oauth_token and key as query parameters, and on 1 November a generateContent request with no credential answered 403 'Method doesn't allow unregistered callers' while one with a made-up ?access_token= answered 401 'Request had invalid authentication credentials' (ACCESS_TOKEN_TYPE_UNSUPPORTED), the deduction Google APIs took on 17 October and 1 November and Gemini Embedding and Gemini Live carry (15 of 30). Terms (effective 23 March 2026, re-read 1 November) say free-tier prompts and responses are used to improve Google products and human reviewers may read them, paid tier isn't used (10 of 20). Abuse monitoring on paid services is kept 'a limited period' with no number, grounding data 30 days, and since 26 October new paid projects store Generate Content requests by default unless store=false. No zero-retention route on this API was found (5 of 15). Request logs in AI Studio for billed projects (15). security.txt on google.com valid to 2030-04-01 with a vulnerability reward programme, re-read on 1 November. No certification naming the Developer API found, the reading Gemini Embedding takes (5 of 20).

Amazon Bedrock 80

AWS Health Dashboard with per-service, per-Region feeds and a public history file (20). The history lists Amazon Bedrock as disrupted in ME-CENTRAL-1 (UAE) and ME-SOUTH-1 (Bahrain) since March 2026, where AWS said on 15 September 2026 (still inside the 90-day window on 6 November) that it can't restore resources held only there. No other Bedrock event appears in the history in the 90 days to 6 November; an EU-SOUTH-2 packet-loss event on 4 October named Bedrock AgentCore, not Bedrock. This counts as one major event (10). StatusGator's third-party mirror shows 'Increased Error Rates' warnings on 24, 26 and 27 August and 10 September that we couldn't find in AWS's own history, so we didn't score them. Per-model token quotas with default values in the AWS General Reference, such as 30,000,000 tokens a minute for Opus 5.5 global inference (15). The error guide gives exponential backoff with jitter for 429, 503 and 529 and says to honour Retry-After, and the SDKs retry ModelNotReadyException up to five times (15). SLA of 99.9 per cent a month per Region for the model APIs, last updated 4 October 2023 (10). Converse and the current Claude, GPT-6 and Nova 2 Lite models are GA (10).

Gemini Developer API 45

Status page at aistudio.google.com/status (robots.txt allows it), but on 1 November it still returned a script shell titled Google AI Studio with no components or history, so 10 of 20, and 5 of 30 for the incident line with no readable record, as the Gemini Live and Gemini Embedding listings read the same page. Google Cloud's status dashboard covers Vertex AI products, not this endpoint. Per-model RPM and TPM live in AI Studio. The public rate-limits page, re-read on 1 November, publishes tier rules, spend-based limits of $10, $50 and $200 per 10 minutes for Tiers 1 to 3, and batch queue limits (5 of 15). The troubleshooting page, re-read on 1 November, gives exponential backoff with jitter for 429 RESOURCE_EXHAUSTED and 503 UNAVAILABLE, says to retry 429, 408 and 5xx and not 400, 402 or 403, and says the Python SDK retries transient errors up to four times (15). No SLA found for the Developer API (0). The flagship gemini-4-argon went GA on 13 October 2026 and the default 3.8 Flash is GA, and the Interactions API the guides lead with has been Generally Available since June 2026. The listing's example and the SDKs default to the v1beta path, which Google labels beta, but the API versions page, read on 1 November, says the stable v1 path carries the Interactions API, generateContent, function calling and structured output and supports every model, so the surface is available as GA, the reading Gemini Embedding takes on the same endpoint. Gemini Live keeps a deduction because the Live API runs on v1beta only (10). Before 13 October the only Pro model was gemini-3.1-pro-preview (5 of 10).

Amazon Bedrock 85

Closed service under the AWS Customer Agreement and Service Terms, with the SDKs and Smithy models under Apache-2.0 (18, as for S3). The data retention, abuse detection and security pages agree that Bedrock stores nothing by default, name the models that retain traffic and for how long, and say retained data isn't shared with providers. The retention page's default mode says AWS 'may retain the data for safety and abuse-prevention purposes' without a period, and a legacy provider_data_share mode is kept though sharing isn't supported (25 of 30). A written lifecycle policy with Legacy and EOL dates per model, but the legacy table's 34-day Legacy periods for five Claude models in some Regions contradict its own 6-month floor (15 of 20, down from 18). AWS's sub-processor list, per the 4 October S3 check, and per-Region, geographic and global inference options, with global routing able to process data in any commercial Region (18 of 20).

Gemini Developer API 72

Closed service with clear terms, SDKs Apache-2.0 (15). Terms, pricing page and logs policy agree that free-tier data improves Google products and paid data doesn't, but abuse retention on paid services has no number, and the 26 October changelog entry turns on request storage by default for new paid projects while existing projects keep their setting (20 of 30). Deprecations table with earliest shutdown dates, exact dates promised later (15 of 20). Terms say data may be cached in any country where Google has facilities. No subprocessor list checked (5 of 20).

Amazon Bedrock 30

No x402, MPP or L402 (0). Per-token prices per model, tier and Region on the public pricing page, filled from AWS's public price feed (20). New accounts get up to $200 in Free Tier credits and the Free Tier page lists Bedrock among eligible services, and the Free Tier FAQ says most new customers don't need a payment method, but AWS may still ask for one, so half, as for S3 (10). A person signs up for AWS in a browser and sets up IAM before any key exists (0).

Gemini Developer API 40

No machine payment protocol (0). Per-token prices published without a login, re-read on 1 November (20). Free tier on 3.8 Flash, 3.6 Flash, 3.5 Flash-Lite and 3.1 Flash-Lite, not on gemini-4-argon or 3.1 Pro preview, and the rate-limits page needs only an active project for the free tier and a linked billing account from Tier 1, so the free tier needs no card (20). A Google account and AI Studio are needed for a key (0).

Amazon Bedrock 94

Model reading. No OpenAPI, but the bedrock-runtime and bedrock Smithy models are public in aws/api-models-aws with types, required members and enums, counted as for S3 (25). llms.txt for the user guide with about 2,000 lines, each page with a Markdown twin (10). Model cards state context, output limits, endpoints and supported functions per model, and the guide now recommends bedrock-runtime for new work, but support for each API, endpoint and feature changes model by model (16 of 20). Structured outputs compile JSON Schema to a grammar, with minimum, maximum and string lengths supported, though some inference profiles don't support it, such as Haiku 4.5 through the India profile (13 of 15). Every operation lists typed errors with HTTP codes and links to a troubleshooting page with causes and fixes, and examples run through the guide (15). API version 2023-09-30 and a dated document history with an RSS feed, newest entry 30 October 2026, with structured outputs for Claude on both endpoints added on 18 October (15).

Gemini Developer API 85

A keyless Discovery document at generativelanguage.googleapis.com/$discovery/rest?version=v1beta (revision 20261101, re-read on 1 November) types 86 methods and 225 schemas for GenerateContent, files, batches, caching and models, but not the Interactions API the guides now lead with (20 of 25). llms.txt at ai.google.dev/gemini-api/docs/llms.txt, with Markdown twins of each page (10). Reference not read in full, and the changelog and latest-model guide name the recommended models (15 of 20). Structured output accepts JSON Schema with enums, minimum, maximum, prefixItems and $ref, but the server ignores minLength, maxLength, minProperties, maxProperties and uniqueItems, per the structured output page re-read on 1 November (10 of 15). Examples on every guide and an API errors page with HTTP, blocked and content codes (15). Dated changelog and model versions (15).