Apodex-1.1-mini

# Fix list: Apodex-1.1-mini

From Anchor Terminal's listing at https://www.anchorterminal.com/tools/apodex-1-1-mini, 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 Apodex-1.1-mini: 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, 65 out of 100, up to 10.8 more on the total

Why it scored 65: File safety 25, safetensors and GGUF only and nothing from the repositories runs. Provenance 15, the files are in the apodex organisation its site links to, with a SHA-256 for every weight file and a SHA256SUMS file for the GGUF. Safety evaluation 0, none published. Agent guidance 5, the recommended harness documents approval gates and a sandbox boundary, but nothing on prompt injection or refusals for the model. Licence risk 15, Apache 2.0 from Qwen3.5-35B-A3B through to this model with no conditions added. Security programme 5, the harness has a disclosure policy with a private address, with no route shown for the weights.

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. Documentation and upkeep, 38 out of 100, up to 9.5 more on the total

Why it scored 38: Model card 8, it covers intended use, context and languages but not the architecture, parameters, limitations or a knowledge cut-off. Training data 5, a general description with no cut-off. Last change 15, the GGUF build went up on 10 September 2026, 41 days ago. Responsiveness 0, Apodex answered none of the four discussions. Version policy 5, 1.0-mini stays in its own repository without a note. File health 5, the files agree, but the card's agent example is broken and reports wait without a reply.

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.

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

Why it scored 85: Memory 20, Apodex's Q4_K_M GGUF with its vision projector is 22.6 GB and needs about 27.1 GB with the fifth added, so 32 GB, as do its NVFP4 and GPTQ-Int4 builds. Runtimes 20, the Qwen3.5 MoE architecture runs in released vLLM, SGLang, Transformers and llama.cpp, while Ollama's and LM Studio's entries are the base's, not this model's. Builds 15, Apodex publishes FP8, NVFP4, GPTQ-Int4 and GGUF. Terms 15, Apache 2.0 with nothing added. Download 10, nothing is gated. Hosted 5, Apodex's own API 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.

## 4. Agent interface, 85 out of 100, up to 2.3 more on the total

Why it scored 85: Chat template 20, a Jinja file in the repository. Tool-calling format 25, the card says how tools go in and come back and names the parser for SGLang and vLLM, with a full example. Trained tool calling 20, trained for native function calling with parallel calls. Structured output 0. Reasoning control 10, the template turns thinking off per request and the card says the reasoning comes back in 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.

## 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.

- Apodex's API serves apodex-1.1-mini, and no Apodex API listing exists in data/tools, so there is no sameModel to link. If one is added, link it
- Security programme rests on FrontierAgent's SECURITY.md, which covers the harness, not the weights. GitHub's advisory list for ApodexAI couldn't be read
- Agent guidance 5 rests on the harness's documented safeguards, read as general cautions
- Whether stock llama.cpp loads the GGUF's multi-token prediction tensors wasn't tested, since nothing was run. The card says a recent release

## Weaknesses

- No safety results, and agent guidance only from the harness
- Apodex answered none of the four discussions
- The card states no parameters, architecture, limitations or knowledge cut-off
- The template adds an Apodex identity block when no system prompt is sent and keeps earlier reasoning by default
- Structured output isn't documented

## 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 it with `vllm serve apodex/Apodex-1.1-mini --enable-auto-tool-choice --tool-call-parser qwen3_coder --reasoning-parser qwen3`, or the same parsers in SGLang, so tool calls and reasoning come back parsed
- For llama.cpp, `llama-server -hf apodex/Apodex-1.1-mini-GGUF:Q4_K_M --jinja --reasoning-format deepseek`, and the card passes the repository's chat_template.jinja with --chat-template-file
- Pass tools through the API's tools parameter rather than in the system prompt, as the card asks
- The card recommends temperature 1.0, top_p 0.95, repetition_penalty 1.05 and max_tokens up to 32,768
- Send your own system prompt, or set identity_mode, since the template otherwise adds an Apodex identity block
- The template keeps all earlier reasoning in the context unless `preserve_thinking` is false, which grows long agent sessions
- Thinking is on by default, send `chat_template_kwargs` with `enable_thinking` false for short tasks, a switch the card doesn't mention
- The card's Python example has a broken line where it runs the tool (available_toolsname), so fix it to `available_tools[name](**args)` before running
- Validate JSON output yourself, since structured output isn't documented

## What the review panel asked for

- fix the agent example
- answer discussions
- a build for 24 GB cards
- name the thinking switch on the card
- a reporting route for the weights
- publish safety results

## 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.