Releases

The changelog lists every change in every release; Upgrading is the shorter list of the ones that change what a deployment serves.

Figure · Releases

The last 8 of 18 releases, newest first.

1. 0.11.12026-10-10
2. 0.11.02026-10-09
3. 0.10.02026-10-08
4. 0.9.02026-10-07
5. 0.8.22026-10-07
6. 0.8.12026-10-05
7. 0.8.02026-10-05
8. 0.7.22026-10-05

Each links to its entry in the changelog below.

Source: CHANGELOG.md at the pinned release.

TL;DR

To see which release you have, ask the installed package, from the decisio checkout the server runs from:

uv run python -c "from importlib.metadata import version; print(version('decisio'))"
# the release number, such as 0.11.1

Run against the CPU stand-in at v0.11.1 on 2026-10-14.

Word	Meaning
release	a tagged version of decisio, with its images on ghcr.io/aminry/decisio and its changelog entry
changelog	CHANGELOG.md, every change in every release, newest first
pinned	the release these pages are built from, named in the footer

The changelog

0.11.1 (2026-10-10)

Bug Fixes

• serve: a request records how long it waited for the engine, and whether a registration was running (#135) (76d484c)
• serve: gemma-4-12b registers a state's boundary before the question, as the 31B does (#128) (b7af9c6)

0.11.0 (2026-10-09)

Features

• hub: the model repositories are under tachara-ai (Tachara AI Lab), pinned at their 0.11.0 revisions (#121) (852cce4)
• serve: --backend mlx serves gemma-4-31b from an MLX conversion, documented for states up to 16,383 tokens (#116) (a801341)

Bug Fixes

• serve: --backend mlx refuses any checkpoint whose config.json declares FP8 quantization (#120) (18229d2)
• serve: a prompt over the context is a 422 with a clear detail, not a bare 500 (#124) (a51be6d)

0.10.0 (2026-10-08)

⚠ BREAKING CHANGES

• serve: a vLLM server started with neither --base nor --model, and the container, now serve gemma-4-31b (a 96 GB card); set --base qwen3.6-35b-a3b or DECISIO_BASE=qwen3.6-35b-a3b to keep the previous Qwen behaviour (docs/upgrading.md).

Features

• hub: the 31B's FP8 checkpoint is made on a CPU as vLLM quantizes it on load, and the engine's FP8 weights can be fingerprinted (#113) (94d675c)
• serve: --base takes a decisio model repository; decision_config.json is exported from each base's profile (#110) (9d352de)
• serve: the default base is Gemma 4 31B, served from its FP8 repository (#114) (389fe13)
• serve: the MLX prefix cache budget is the base's: 7,400 MiB for gemma-4-12b, 2,048 MiB for the Qwen base (#112) (71facc7)

0.9.0 (2026-10-07)

Features

• serve: register a state's boundary after the response on the Gemma bases, so a first question does not wait for it (#99) (787df0e)

0.8.2 (2026-10-07)

Bug Fixes

• serve: --backend mlx refuses gemma-4-31b by the resolved base, so a 31B conversion without --base is refused too (#100) (f6cc1e8)

0.8.1 (2026-10-05)

Bug Fixes

• serve: a single question registers its state's boundary on the Gemma bases, so the next question reads the state from the cache (#93) (573d7d6)
• serve: close the engine-death races; loop those tests under load nightly with the slow tier on a hosted runner (#91) (3c51e79)

0.8.0 (2026-10-05)

Features

• serve: the engine runs in the server's process by default; gemma-4-31b scores several questions warm (#86) (efd2b83)

0.7.2 (2026-10-05)

Bug Fixes

• serve: a dead engine turns /health to 503, is refused fast and ends the server with code 70 (#78) (8b367d9)

0.7.1 (2026-10-05)

Bug Fixes

• serve: --head-engine gives the second engine the server's prompt format and base (#76) (993db22)

0.7.0 (2026-10-05)

Features

• serve: --engine-process in|separate, where vLLM's engine runs (default separate) (#74) (bbff50f)

0.6.0 (2026-10-05)

Features

• serve: Gemma 4 31B as a third base (--base gemma-4-31b), FP8 on load (#71) (73f0eff)

0.5.0 (2026-10-04)

Features

• the Gemma base on a Mac with --backend mlx, 6-bit by default (#69) (587c8b2)

0.4.1 (2026-10-04)

Bug Fixes

• serve: load the pinned revision in packed mode, the hf stand-in and the second-engine head (#65) (b0b4ea5)
• serve: pin the Qwen base's checkpoint revision at 95a723d (#63) (9a36b96)

0.4.0 (2026-10-04)

Features

• serve: Gemma 4 12B as a second base (--base gemma-4-12b) (#48) (61d7519)

Bug Fixes

• conformance: C4 applies each question type's served temperature (#60) (4ab98af)

0.3.0 (2026-10-04)

⚠ BREAKING CHANGES

• serve: tasks registered under the earlier compact layout are not applied by the new default (the layout enters the task fingerprint and the yes/no rendering a yes/no question's task key); register them again, or serve them with --prompt-tail compact --noul-rendering words. --mode packed reads only the compact layout and needs --prompt-tail compact.

Features

• demos: four demos on a System One server, judged decision by decision, with clips rendered from trajectories (#33) (86abd52)
• serve: --noul-commit, an opt-in transform reporting yes/no answers inside 0.20 to 0.80 at the band's edge (#53) (62dfe1f)
• serve: the spaced layout and yes/no as named letters are the served default, with T 1.506 and choice T 1.370 (6c581d4)

0.2.0 (2026-10-03)

⚠ BREAKING CHANGES

• serve: no padding (--pad-policy none) is the MLX backend's default (#46)
• --multi-question sequential and --describe-options are the served defaults (#38)

Features

• --describe-options shows an option's description alone when it has one (off by default) (#36) (f2d936e)
• --multi-question sequential and --describe-options are the served defaults (#38) (06a1273)
• --multi-question sequential makes each answer equal the question asked alone; stage timings and --pad-policy; README wording for multi-question requests (#35) (5461e0e)
• serve: --noul-rendering words|letters|letters-keys (words by default) (#44) (597f383)
• serve: --noul-rendering words|letters|letters-keys, how a yes/no question is asked (words by default) (597f383)
• serve: --pad-policy row, a single-question row padded to end on the block boundary (off by default) (#45) (8ccf143)
• serve: a cross-request prefix cache for the MLX engine (--prefix-cache-mb) (#47) (8e3f54f)
• serve: an MLX engine for the text route on Apple silicon (--backend mlx) (#39) (f3c5026)
• serve: no padding (--pad-policy none) is the MLX backend's default (#46) (2c1204e)

Bug Fixes

• task registration on a certain readout, and the CPU stand-in's head engine (two HTTP 500s) (#49) (6e71298)

0.1.1 (2026-10-01)

Bug Fixes

• ci: attest a CycloneDX SBOM of the image instead of SPDX (#22) (eef30b4)

0.1.0 (2026-10-01)

Features

• a generator for the conformance items file from two public sets (#18) (77f1bbd)

Bug Fixes

• the head's output-layer rows load from a Hugging Face repo id as well as a local directory (9d7d0ab)

Documentation

• community files, CI and release tooling (#2) (4dc5388)
• launch checklist for the public flip (#4) (c79ec96)
• launch checklist gate for the image's first GPU start; README says it is pending (dff92fb)
• launch checklist records the flip and corrects three steps (#14) (93a885d)
• launch checklist section for the maintainer's commit signing (80e36ca)
• README section Run with Docker; launch checklist covers the image (2e93c65)
• record that the merge settings of checklist 2.1 are applied (7b60810)
• record the history rebuild of main in the launch checklist (b5688cc)

What next

• Upgrading: what changes for a running deployment between releases.
• Choosing a base: what the default base is, and when it changed.
• The evaluation card: what each release was measured on.