occamy-1.0

# Fix list: occamy-1.0

From Anchor Terminal's listing at https://www.anchorterminal.com/tools/occamy-1-0, the October 2026 research run, assessed 21 October 2026. Grade C, 61 out of 100.

This is everything the published grade says the listing lacks, the biggest possible gain to the total first. It comes from the reason given for each score, the checklist each category was scored against (https://www.anchorterminal.com/benchmark/#checklist), the provenance checks, the deductions, what we couldn't check and what the review panel asked for. A fix counts at the next check, once it's public.

For a coding agent working on occamy-1.0: work through the items below in the product, its docs and its public pages. Each category gives the reason for its score, with the points each checklist item earned, and the checklist itself, so the gap is the items that earned less than their points. Change the product, not the wording, and keep a note of what you changed and where it's published.

## 1. Risk, 55 out of 100, up to 13.8 more on the total

Why it scored 55: File safety 25, safetensors only and nothing from the repository runs to load it or build prompts. Provenance 15, the files are in the organisation Accio's GitHub links, with MD5 sums and SHA-256 for every file. Safety evaluation 0, none published. Agent guidance 0. Licence risk 15, plain Apache 2.0 that hasn't changed. Security programme 0, bug reports go to public GitHub issues and there's no security contact.

The checklist (https://www.anchorterminal.com/benchmark/#checklist-risk):

What could go wrong when you run it, and what the publisher has done about it. A higher score means less risk.

- 0 to 25, file safety. 25 for weights in safetensors or GGUF only, 15 when pickle files (.bin, .pt, .pth) sit beside them, 0 for pickle only. 10 off when the documented way to load it or to build its prompts runs code from the repository (`trust_remote_code`, or the publisher's scripts in place of a template).
- 0 to 15, provenance. 10 when the files come from the publisher's own organisation, the one its site or docs link to, and 5 for a checksum on every file.
- 0 to 20, safety evaluation. 20 when results are published and someone other than the publisher ran them or can rerun them, 10 for the publisher's own results, 0 for none.
- 0 to 15, guidance for agent use. 15 for documented prompt-injection guidance or an instruction hierarchy the model was trained to follow, 10 for a statement of what its trained refusals cover and the safeguards to add around it, 5 for general cautions, 0 for none.
- 0 to 15, licence risk. 15 to start, 5 off for each of these, a use policy or field-of-use limits beyond the law, terms the licensor can change or revoke, a revenue or user threshold, and a licence that changed without a dated note or whose link is broken. Not below 0.
- 0 to 10, a security programme. 5 for a way to report a flaw (security.txt, a disclosure policy or a security contact), 5 for advisories handled in public.

When our reference-machine runs start, measured robustness (prompt injection with AgentDojo, over-refusal with XSTest) joins Risk.

## 2. Reach, 85 out of 100, up to 5.8 more on the total

Why it scored 85: Memory 25 of 30, Accio's MLX mxfp4 build is 18.4 GB, 22.1 GB with a fifth added, and its IQ4_XS GGUF needs 22.5 GB, both within 24 GB. Runtimes 20, it keeps the base's qwen3_5_moe architecture, which llama.cpp, vLLM, SGLang and Transformers run in released versions, while Ollama's entry is the base's. Builds 15, FP8, NVFP4, GGUF and MLX from Accio. Terms 15, Apache 2.0 like its base. Download 10, nothing is gated. Hosted 0, no API or inference provider serves it.

The checklist (https://www.anchorterminal.com/benchmark/#checklist-reach):

Whether an agent can run it, and where.

- 0 to 30, memory. The smallest build from the publisher or a runtime's own library that keeps 4-bit precision or better, at its file size plus a fifth for the runtime and a working context. 30 for 16 GB or less, 25 for 24 GB, 20 for 32 GB, 15 for 64 GB, 10 for what fits our reference machine (the GPU share of 128 GB of unified memory), 5 for 256 GB, 0 above.
- 0 to 25, runtimes. 5 for each of llama.cpp, Ollama, vLLM, SGLang, LM Studio and Hugging Face Transformers that runs the architecture in a released version, from the runtime's own code, library or documentation. A publisher's fork or an unreleased branch doesn't count. 25 at most.
- 0 to 15, builds from the publisher. 15 for quantised builds at two precisions or more (FP8, 4-bit, GGUF), 10 for one, 0 when only other people's quantisations exist.
- 0 to 15, terms of use. 15 for an OSI-approved licence with nothing added, 10 for a custom licence that lets anyone use it commercially for free below a threshold or under a use policy, 5 for research or non-commercial use only, 0 for no licence or unclear terms.
- 0 to 10, download. 10 when the files download with no account and no terms to accept, 5 for a click-through that needs an account, 0 when a person approves each request.
- 5, hosted too. The publisher's own API serves it, or at least two inference providers do, so an agent without the hardware can still call it.

When our reference-machine runs start, speed on that machine (tokens per second and time to first token at the build that fits) joins Reach.

## 3. Agent interface, 70 out of 100, up to 4.6 more on the total

Why it scored 70: Chat template 20, the base's Jinja template in the repository. Tool-calling format 10 of 25, the card names the parser but the format lives only in the template. Trained tool calling 20, trained on tool use with several calls a turn. Structured output 0, not documented. Reasoning control 10, enable_thinking per request and the card says to keep the returned reasoning content. Context 10, 262,144 tokens.

The checklist (https://www.anchorterminal.com/benchmark/#checklist-interface):

What a model and a developer need to talk to it. Speech, image and other models with no chat interface read the tool-calling and structured-output lines for what the model takes and returns (voices, timestamps, streaming, typed outputs), and the note says how.

- 0 to 20, the chat template. 20 for a template in the repository (Jinja, in the tokenizer config or its own file), 10 when the prompt format is only given as code you run.
- 0 to 25, the tool-calling format. 25 when the publisher documents how tools are declared and how calls come back, with the parser to use in a named runtime, 10 when only the template or a runtime's own recipe says it.
- 0 to 20, tool calling the model was trained for. 20 with several calls in one turn, 10 for one call at a time.
- 0 to 15, structured output. 15 when JSON or JSON Schema output is documented for the model with how to ask for it, 5 when the card claims it without saying how.
- 10, reasoning control. Thinking that can be turned off or budgeted per request, with where the reasoning text comes back.
- 0 to 10, context. 10 for 128K tokens or more trained, 5 for 32K to 127K. Context reached only by scaling the position encoding counts one band down.

## 4. Documentation and upkeep, 74 out of 100, up to 4 more on the total

Why it scored 74: Model card 14 of 20, languages and a knowledge cut-off are missing. Training data 15, the mix by type with token counts, the method and a published subset, without a cut-off. Last change 20, the card changed on 2 October 2026. Responsiveness 10, Accio answered 4 of 9 questions. Version policy 5, a first release with no files replaced. File health 10, the files agree with the card.

The checklist (https://www.anchorterminal.com/benchmark/#checklist-upkeep):

- 0 to 20, the model card. Intended use, limitations, architecture, parameters, context length, languages and knowledge cut-off. 20 for all seven, 3 off for each that's missing.
- 0 to 20, training data. 20 for the data published or listed with sources and sizes, 15 for a data card that names the sources or the mix, the cut-off and the filtering, 10 for a general description with a cut-off, 5 without one, 0 for none.
- 0 to 20, the last change to the files, the template or the reference code. 20 within 30 days, 15 within 90, 10 within 180, 5 within a year, 0 older.
- 0 to 20, responsiveness. The publisher's replies to the 40 most recent questions and reports in the repository's discussions or issues. 20 for most, 10 for some, 5 for a few, 0 for none.
- 0 to 10, version policy. 10 for earlier versions kept in their own repositories with a dated note when one is replaced, 5 for earlier versions kept without notes or a first release, 0 when files were replaced in place without a note.
- 0 to 10, file health. 10 when config, tokenizer, template and card agree and no report of a broken file waits without a reply, 5 when one does, 0 when the files contradict the card.

## What we couldn't check

What we couldn't read counted as absent. Publishing it on a page a plain HTTP fetch can read (not only in a browser) lets the next check count it.

- Training data was read as 15, between the 20 line (data published or listed with sources and sizes) and the settled rule that a description without a cut-off scores 5. A strict cut-off reading gives 5 and upkeep 64
- Who Accio is. The report's byline is Accio Team, the Hugging Face organisation isn't verified and no legal entity is named. Two discussion titles ask whether it's part of Alibaba; we didn't read them and found no first-hand statement
- unchecked, GitHub issues on Accio-Lab/occamy, since the GitHub API wasn't available to this session
- The card's licence line says to see the Hugging Face card for the weights' terms, which is the same card. The metadata and LICENSE both say Apache 2.0, so no deduction
- The published data subset carries no licence on its card

## Weaknesses

- No safety results, prompt-injection guidance or security contact
- No hosted API or inference provider
- Tool-calling format and structured output aren't documented beyond the template and parser name
- Languages and a knowledge cut-off aren't stated
- The MLX builds await Apple Metal acceptance, per Accio's own cards

## What costs an agent a turn today

The notes we give agents before they call it. Each one is a workaround an agent shouldn't need.

- Serve with `--reasoning-parser qwen3 --enable-auto-tool-choice --tool-call-parser qwen3_coder` on vLLM 0.19.0 or SGLang 0.5.10 and later
- BF16 is 70.2 GB, so on a 128 GB machine run it without the card's 8-way tensor parallelism
- Keep the whole assistant message, reasoning and tool calls included, in later turns, as the card says
- Sampling per the card is temperature 1.0, top_p 0.95, top_k 20 and presence_penalty 1.5
- With the GGUFs, NFC-normalise input text and read TOKENIZER.md, since the files store qwen2 pre-tokenizer metadata
- The base's MTP layer isn't in this checkpoint, Accio-Lab/occamy-1.0-MTP is an experimental separate head

## What the review panel asked for

- name the publisher
- a hosted endpoint
- describe returned calls on the card
- a private security contact

## When it's done

Send what changed and where it's published as a dispute (https://www.anchorterminal.com/builders/#disputes, or `POST https://www.anchorterminal.com/api/v1/contact` with `"kind": "dispute"`). Disputes are answered in public, and the listing is checked again by the same checklist. Paying for an audit or a listing claim changes nothing here.