Recursive Flint

# Fix list: Recursive Flint

From Anchor Terminal's listing at https://www.anchorterminal.com/tools/recursive-flint, the October 2026 research run, assessed 22 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 Recursive Flint: 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, 60 out of 100, up to 12.3 more on the total

Why it scored 60: File safety 15 of 25, safetensors and GGUF only, but the card's documented BF16 route installs the wheel shipped in the repository. Provenance 15, the files are in the RecursiveSI organisation that Recursive's own engine downloads from, with a checksum on every file. Safety evaluation 0, none found, and the announcement post couldn't be read. Agent guidance 10, the card and serving README say to enforce permissions around tool use and to strip tool tags from untrusted tool output, but there's no trained instruction hierarchy. Licence risk 15, plain Apache 2.0. Security programme 5, a reporting address on the card and a private reporting route for the engine, with no public advisories yet.

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, 75 out of 100, up to 9.6 more on the total

Why it scored 75: Memory 20 of 30, Recursive's MLX INT4 build is 22.6 GB, or 27.1 GB with a fifth added, which fits 32 GB, and every other 4-bit build lands in the same band. Runtimes 15 of 25, released llama.cpp (b11803), Ollama 0.40.8 with its library entry and LM Studio's catalogue run it, while SGLang needs Recursive's plugin, Transformers support isn't released and vLLM has none. Builds 15, MLX INT4 plus Q4_1 and NVFP4 GGUFs from Recursive. Terms 15, Apache 2.0 with nothing added. Download 10, nothing is gated. Hosted 0, no inference provider serves it and no Recursive API was confirmed.

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. Documentation and upkeep, 67 out of 100, up to 5.1 more on the total

Why it scored 67: Model card 17 of 20, everything but languages. Training data 10, a general description of the mix with a June 2026 cut-off. Last change 20, released one day ago. Responsiveness 5, Recursive answered one of the four first-day discussions. Version policy 5, a first release. File health 10, the files agree with each other and the card, and no broken-file report waits.

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.

## 4. Agent interface, 90 out of 100, up to 1.5 more on the total

Why it scored 90: Chat template 20, a Jinja file in the repository and inside both GGUFs. Tool-calling format 25, Recursive's serving README documents how tools go in, how calls come back and the flint parser. Trained tool calling 20, several calls a turn. Structured output 15, response_format with JSON Schema is documented. Reasoning control 0, the model decides when to reason and Recursive says there's no switch or budget. Context 10, 262,144 tokens with no position 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.

- unchecked, Recursive's announcement post and docs (www.recursive.com, docs.recursive.com), which our reader couldn't open on 22 October 2026, so a hosted Recursive API, published safety results and security.txt are scored as absent
- unchecked, rec's GitHub issues, so responsiveness rests on the four Hugging Face discussions
- Runtimes credits Ollama for its MLX runner in 0.40.8; whether its GGUF tags run on Linux in a released Ollama wasn't confirmed, since its bundled llama.cpp (b11738) predates Flint
- SGLang support is Recursive's flint_sglang plugin, scored 0 as the second batch read Spark X2.6's vLLM plugin; the first batch credited Kolibri's vLLM plugin 5, so the two readings differ
- Transformers and SGLang support may arrive in their next releases, which would add 10 to runtimes
- Capability isn't scored here, and no independent result exists yet

## Weaknesses

- No independent benchmark results yet, every published score is Recursive's
- No reasoning switch or budget, so hard prompts can spend tens of thousands of tokens thinking
- The BF16 route needs one 141 GB GPU and Recursive's SGLang plugin, and vLLM and released Transformers don't run it
- No safety results and no trained instruction hierarchy
- No inference provider or confirmed Recursive API for agents without the hardware

## 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 `ollama run flint` (Ollama 0.40.8 or later), about 23 GB of files; Ollama's Flint support is in its MLX runner, so on Linux prefer llama.cpp
- With llama.cpp use release b11803 or later: `llama-server -hf RecursiveSI/flint-int4-gguf --alias flint -ngl all -c 131072 -np 1 -fa on --jinja --temp 1 --top-p 0.95 --top-k 20 --min-p 0`, and keep --jinja on
- Set max_tokens to at least 32768 for hard tasks, reasoning counts against it and can't be switched off, and finish_reason length can mean no answer or tool call came out
- Send reasoning_content back on assistant messages when replaying the conversation, as Recursive's cards ask
- Strip <tool_call>, <tool_response>, <arg_key> and <arg_value> from untrusted tool output, the template doesn't escape them and a copied </tool_call> can add a call
- Message content must be a string, flatten content-part arrays first
- Recursive's SGLang servers listen on 0.0.0.0 with no key, publish the port on localhost only or pass --api-key

## What the review panel asked for

- publish safety results
- bind localhost by default
- publish dated release notes
- upstream the SGLang support
- a reasoning budget or switch
- memory figures per context length

## 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.