# Fix list: Spark-X2.6-4B From Anchor Terminal's listing at https://www.anchorterminal.com/tools/spark-x2-6-4b, 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 Spark-X2.6-4B: 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, 50 out of 100, up to 15.4 more on the total Why it scored 50: File safety 15 of 25, safetensors and GGUF only, but the card's Transformers and vLLM routes run the repository's Python. Provenance 15, the files are in SparkLLM's own Hub organisation, which its GitHub docs name, and every file has a checksum, though neither organisation is verified and no company site was found. Safety evaluation 0 and agent guidance 0, none published. Licence risk 15, plain Apache 2.0 that hasn't changed. Security programme 5 of 10, a private reporting route and email in SparkLLM's community repository, with no public advisories. 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, 41 out of 100, up to 9.1 more on the total Why it scored 41: Model card 11 of 20, intended use, parameters, context and languages, with no limitations, cut-off or architecture description of its own. Training data 0, the X2.6 card describes none. Last change 20, the files went up on 19 and 20 October. Responsiveness 0, nobody has opened a discussion yet. Version policy 5, X2.5 stays in place with no note. File health 5, the tokenizer config's 131,072 disagrees with the 1M context. 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 30, SparkLLM's Q4_K_M GGUF is 3.19 GB, about 3.8 GB with a fifth added, which fits 8 GB. Runtimes 15 of 25, released llama.cpp, SGLang and Ollama run the spark2_5 architecture X2.6 keeps, while vLLM needs SparkLLM's plugin, Transformers needs the repository's code and LM Studio has no catalogue entry. Builds 15, SparkLLM publishes FP8, INT8, INT4 and GGUF builds. Terms 15, Apache 2.0 with nothing added, on its own base. Download 10, nothing is gated. Hosted 0, no provider or SparkLLM 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, 70 out of 100, up to 4.6 more on the total Why it scored 70: Chat template 20, a Jinja file in the repository. Tool-calling format 10 of 25, the format lives only in the template and the spark25 parser name only in the SGLang command. Trained tool calling 20, the template renders several calls in one turn. Structured output 0, not documented. Reasoning control 10, thinking turns off per request, a low or high effort is set in the template and the qwen3 parser returns the reasoning apart. Context 10, 1,048,576 tokens with no scaling. 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. - Model card architecture is scored as missing because the X2.6 card only says it's the same as Spark-X2.5-4B and doesn't link it. A reader who accepts the reference would give 14 and Upkeep 44 - Training data is scored 0 because the X2.6 card describes none. A reader who carries X2.5's general description over would give 5 - unchecked, whether LM Studio's runtime runs spark2_5, as the card says. LM Studio's catalogue has no Spark entry, so it's scored 0 - Responsiveness is 0 only because the repositories are two days old and nobody has asked anything yet. It should be recounted at the next check - The Base repository's card is a cut-down copy of the post-trained card that calls its own files the post-trained model, though its shards differ from both this repository's and X2.5-Base's. Whether it's a true base checkpoint is unconfirmed - No SparkLLM website was found, so provenance rests on the GitHub organisation naming the Hub repositories. Who SparkLLM and XHToken are as a company is unconfirmed ## Weaknesses - No safety results and no agent or prompt-injection guidance - Transformers and vLLM need code from SparkLLM's repositories - Tool format and structured output aren't documented, and the template uses its own arg_key format - The card has no limitations, cut-off, architecture description or training data, and leans on X2.5's - Released on 19 and 20 October 2026, with no downloads, discussions or independent results yet ## 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. - Quickest start is `llama-server -hf XHToken/Spark-X2.6-4B-GGUF:Q4_K_M --jinja`, which needs llama.cpp b10828 or later, or `ollama run SparkLLM/Spark-X2.6-4B` (8.2 GB, Ollama 0.34.1 or later) - For parsed tool calls serve it with SGLang 0.5.22 or later, `--tool-call-parser spark25 --reasoning-parser qwen3`, since the calls come back as tool_call blocks with arg_key and arg_value pairs - Thinking is on by default. Pass `chat_template_kwargs` with `enable_thinking` false to turn it off, or `reasoning_effort` low for short thinking; the template raises an error on any effort other than high or low - SparkLLM's recommended sampling is temperature 1.0, top_p 0.95 and top_k -1 - The template always opens with its own default system line, with the effort instruction, and puts your system message after the tool list - The template raises an error if tool_call arguments are a JSON string, so pass them as an object - Set the context yourself, the card's 1,048,576-token example needs far more memory than the weights ## What the review panel asked for - add a successor note to X2.5 - correct model_max_length - a quantised Ollama tag - describe the arg_key format - native runtime support - 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.