Gemini Decision 1 vs Granite Switch 4.1

Gemini Decision 1 60

Model reading for gemini-decision-1 on the Gemini API's Decisions endpoint. Platform facts are the Gemini API listing's (gemini-api, re-read 1 November), carried over as RULES.md allows for a model on an already graded platform. No machine payment protocol (0). The pricing page lists $0.10 per million input tokens on the paid tier for every modality and no output charge, without a login (20). Free of charge on the free tier (20). A Google account and an AI Studio key come first (0). 0 + 20 + 20 + 0 = 40. Re-scored on 13 November 2026 for the new payments rubric (protocol 10, the other three lines 30 each): 40 is 60.

Granite Switch 4.1 90

Self-hosted rule. Nothing to buy, and the Apache 2.0 licence allows commercial use at no charge (20). Free to run with no card (20). The weights are not gated, so an agent can fetch and run them with no sign-up (20). No payment protocol (0). IBM hosts no endpoint for Granite Switch that we found, and Hugging Face lists no inference provider for it. Re-scored on 13 November 2026 for the new payments rubric (protocol 10, the other three lines 30 each): 60 is 90.

Gemini Decision 1 80

Model reading for gemini-decision-1 on the Gemini API's Decisions endpoint. Platform facts are the Gemini API listing's (gemini-api, re-read 1 November), carried over as RULES.md allows for a model on an already graded platform. No published machine-readable contract for the Decisions endpoint; the request and response types are generated and typed in google-genai 3.2.0, Apache-2.0 (10 of 25). llms.txt with Markdown twins on the platform (10). The guide and model page say what the endpoint is for and list what it doesn't do (no caching, Batch, Flex, Priority, function calling, structured outputs or text generation; no YouTube or gs:// sources; history only as text) (18 of 20). Fields carry stated rules: question keys match [A-Za-z0-9_-]+ up to 64 characters, 1 to 1,000 questions, 2 to 255 options or levels, option keys 1 to 128 characters, P(true) between 0.01 and 0.99, nesting to 100 levels (14 of 15). Python, JavaScript and REST examples, and errors documented as HTTP 400 with codes safety, prohibited_content, blocklist, image_safety and content_blocked (13 of 15). Dated SDK changelog and the platform's dated changelog (15). 10 + 10 + 18 + 14 + 13 + 15 = 80.

Granite Switch 4.1 57

Read for a model the owner serves. The route is vLLM's OpenAI chat-completions API, and the adapter contract is the chat template, adapter_index.json and one io.yaml per adapter. There is no specification file of its own (10 of 25). We found no llms.txt. The model cards, BUILD.md, the adapter card and the repository docs are Markdown on Hugging Face and GitHub (5 of 10). The card says what each adapter does, that decision probabilities are a prior and not a certified classification outside trained domains, that the guardian adapter is no substitute for application testing, and the speech limits (16 of 20). A decision request is free-form prompt text with token ids the caller computes, and the guardian criterion is free text. The adapter card defines choice, noul and score types, but nothing on the endpoint enforces them (6 of 15). Python and curl examples for both adapters and audio, a chunking recipe for wide questions, and two documented HTTP 400 cases (11 of 15). Versioned model names, and since 27 October a dated What's New section in each card with a revision table and a Hub tag that pins 4.1 (granite-switch-4.1-8b, granite-switch-4.1-30b). The package changelog still has one entry, and the card names a granite-switch-4.1.1-8b tag that the Hub's tag list doesn't have (9 of 15, 7 on 17 October). 10 + 5 + 16 + 6 + 11 + 9 = 57.

Gemini Decision 1 85

Read on the tool lines, as the other decision-model listings are. Model reading for gemini-decision-1 on the Gemini API's Decisions endpoint. Platform facts are the Gemini API listing's (gemini-api, re-read 1 November), carried over as RULES.md allows for a model on an already graded platform. Answers are probabilities only, and the shared input is billed once across up to 1,000 questions (25 of 25). Limits are written down: 262,144 tokens shared by the input and all questions, 1,000 questions, 2 to 255 options, inline data to 100 MB (50 MB for PDFs), 15 MB per external URL (20 of 20). Errors return HTTP 400 with a named code an agent can branch on, plus the platform's backoff advice (15 of 20). SDK retries and backoff guidance, no idempotency key; nothing is written, so a repeat only costs input again (10 of 20). Three required fields, SDK 3.2 or later in Python and JavaScript, and REST (15 of 15). 25 + 20 + 15 + 10 + 15 = 85.

Granite Switch 4.1 66

Read as an endpoint an agent calls for a decision, as with Vela and Decision 3.0. A decision returns one token and its label logprobs, a guardian check up to 16 tokens, and the state turn can share vLLM's prefix cache with chat. Context is 131,072 tokens (21 of 25). The caller picks the adapter per request, the options and the label ids. More than 128 options need one request per 128 (13 of 20). Errors are vLLM's, with the 128-id limit and the audio prompt-length error documented. A guardian answer is JSON in text with no stated handling for a malformed reply (8 of 20). Requests are stateless and the examples set temperature 0, so a decision can be repeated safely (16 of 20). The OpenAI SDK works, and Mellea wraps the guardian adapter, but the decision adapter has no Mellea call yet, label ids need the tokeniser, and vLLM needs a GPU (8 of 15).

Gemini Decision 1 79

Model reading for gemini-decision-1 on the Gemini API's Decisions endpoint. Platform facts are the Gemini API listing's (gemini-api, re-read 1 November), carried over as RULES.md allows for a model on an already graded platform. Released on 11 November 2026, the day of this check (30). Deprecations with earliest dates and no stated minimum notice (4 of 12). Shutdowns in the last 90 days kept to their table (4 of 8). Dated changelogs: google-genai 3.2.0's entry adds the Decisions API; the platform changelog wasn't re-read today (11 of 15). SDK replies not sampled (5 of 10). google-genai 3.2.0 tagged the same day with client.decisions.create; the JavaScript SDK is per Google's post (15). Runtimes stated on the platform (10). 30 + 4 + 4 + 11 + 5 + 15 + 10 = 79. The model is hours old, so these lines are the platform's habits more than this model's record.

Granite Switch 4.1 68

Read as a model, for deprecation notice and model churn rather than release counts, on the lines used for Granite Guardian, Decider, Kev and Strands Decider (recency 30, deprecation 12, churn 8, responsiveness 25, packages 15, health 10). The 17 October note used the tool lines. Revision 4.1.1 of the 8B and 30B checkpoints, with updated decision weights and speech input in seven languages, was published on 27 October 2026, 8 days before the check (30). No deprecation policy. 4.1.1 replaced 4.1 on main with a dated What's New note and a Hub tag that pins 4.1, while the preview repositories were rebuilt in place on 3 May and 15 October 2026 (5 of 12, level with Decider and Granite Guardian). Previews on 1 May, 4.1 on 14 October and 4.1.1 on 27 October, two replacements of main in 13 days with the earlier one still loadable by tag (4 of 8, level with Decider and Strands Decider). The clone's 200 newest commits come from 19 authors, with pull requests numbered to #154 and merged within days, and GitHub issue reply times couldn't be read. On the Hub, IBM merged its own pull requests and answered one of four questions (30B thread 5, six minutes after it was opened on 31 October), while the threads of 18 October (8B), 26 October (30B GGUF) and 2 November (30B) had no reply on 4 November, counted by author and date only (10 of 25). An official Python package with a vLLM plug-in, IBM's own GGUF builds of the 8B and 30B since 22 October (made from 4.1, before 4.1.1), and Mellea for the guardian adapter. The decision adapter still has no Mellea call, and whether llama.cpp runs the composed switch layer is unchecked (11 of 15). CI, a lock file and dependency ranges that track vLLM 0.26 to 0.30 (8 of 10). 30 + 5 + 4 + 10 + 11 + 8 = 68.

Gemini Decision 1 45

Model reading for gemini-decision-1 on the Gemini API's Decisions endpoint. Platform facts are the Gemini API listing's (gemini-api, re-read 1 November), carried over as RULES.md allows for a model on an already graded platform. Keys bound to a service account and restricted to the Gemini API, with no scopes and a token accepted in the URL (15 of 30, the platform's deduction). The pricing page, read on 11 November, says free-tier content is used to improve Google's products and paid-tier content isn't (10 of 20, the platform's reading). Abuse retention with no number (5 of 15). Request logs for billed projects are described for Generate Content; the Decisions guide says nothing on logging (10 of 15). security.txt and a reward programme, no certification naming the API (5 of 20). 15 + 10 + 5 + 10 + 5 = 45.

Granite Switch 4.1 54

Read as a model the owner runs, on the model lines as Granite Guardian is, with retention and training in place of the read-only line. No account, key or gate, so there is no secret to leak, and SECURITY.md tells operators to add authentication to the vLLM server (14 of 30). Self-hosted, so inputs stay on the owner's hardware and nothing reaches IBM or trains a model, but no IBM page says so (15 of 20, level with Granite Guardian and Strands Decider). It reads untrusted text and audio by design and carries a safety adapter with a jailbreaking criterion. The card warns that a mistranscription can silently change what an adapter answers and that the guardian adapter is no substitute for safety testing. Nothing documents how hostile text in the state can steer a decision (7 of 15). No call log of its own beyond vLLM's. Decisions return a probability for every option (5 of 15). SECURITY.md takes private reports through GitHub or the maintainers, IBM's security.txt is valid (now expiring 4 December 2026) with PSIRT and HackerOne, commits need a DCO sign-off, and the decision adapters in Granite Libraries are signed with certificates. Since 23 October 2026 IBM's own GGUF builds of the 8B and 30B carry a model.sig from its signing workflow, but they were made from 4.1, and the composed safetensors checkpoints the card serves with vLLM, 4.1 and 4.1.1 alike, still have no signature file. Granite Guardian's 15 rests on signed safetensors weights, so this stays below it (13 of 20, as on 17 October). 14 + 15 + 7 + 5 + 13 = 54.

Gemini Decision 1 40

Model reading for gemini-decision-1 on the Gemini API's Decisions endpoint. Platform facts are the Gemini API listing's (gemini-api, re-read 1 November), carried over as RULES.md allows for a model on an already graded platform. Status page at aistudio.google.com/status with no readable components or history (10 of 20) and no readable incident record (5 of 30). Rate-limit tiers and spend limits published for the platform, but the public rate-limit page, read on 11 November, has no row for Decision 1 and sends you to AI Studio (5 of 15). Backoff guidance for 429 and 503 and SDK retries (15). No SLA (0). The guide and the model page, read on 11 November, state no launch stage; the endpoint is v1beta and the model has one version, so half, as Gemini 4 Carbon was read (5 of 10). 10 + 5 + 5 + 15 + 0 + 5 = 40.

Granite Switch 4.1 46

Scored on the local-software lines, since the owner serves the checkpoint with vLLM. The README installs with pip install "granite-switch[vllm]", and the model card (revision 4.1.1) now states Python 3.11 to 3.13, PyTorch 2.11 or later, Transformers 5.16 to 5.17.0 and vLLM 0.26 to 0.30, matching pyproject.toml. pypi.org didn't connect on 4 November, and the publish workflow runs on a GitHub release while the remote has no tags, so whether 0.1.1 reached PyPI is unread (12 of 20, held). Public CI runs pre-commit and CPU unit tests on Python 3.11 and 3.12 on every push and pull request, and GPU suites run on a self-hosted runner when a maintainer asks. GitHub's pages answer 403 for our reader, so whether main is passing is unread (15 of 25). Open crash and regression issues on GitHub couldn't be read for the same reason. The Hugging Face repositories have had four questions since 18 October, one answered by IBM (10 of 25, GitHub unchecked). CHANGELOG.md follows Keep a Changelog and semver but still has one entry, 0.1.0 MVP, with none for 0.1.1, and the repository has no tags (4 of 15). The checkpoints carry no preview label, but the composer is 0.1.x and its changelog says the API is not yet stable (5 of 15). 12 + 15 + 10 + 4 + 5 = 46.