From 985231410d97e63926fa777e365b6c1b464ba7e1 Mon Sep 17 00:00:00 2001 From: "J.Jason" <130959319+JJasonSun@users.noreply.github.com> Date: Mon, 19 Oct 2026 13:28:28 +0800 Subject: [PATCH 1/2] docs: sync model facts with 2026-10-01 ECNU update - ecnu-max moved to DeepSeek-V4.1-Flash-1017 with native image understanding - ecnu-plus moved to Qwen3.8-Flash-Next with low/medium/xhigh effort tiers - add error-code pointer, concurrency cap and TTS model facts - correct WorkBuddy text-only rationale and dated contract notes Co-Authored-By: Claude Code --- references/agent_development.md | 17 +++++++++-------- references/api_reference.md | 1 + references/known_deviations.md | 19 +++++++++++-------- references/models.md | 27 +++++++++++++++++++-------- references/workbuddy_setup.md | 11 +++++++---- references/workflows.md | 4 ++-- 6 files changed, 49 insertions(+), 30 deletions(-) diff --git a/references/agent_development.md b/references/agent_development.md index ee96309..28df877 100644 --- a/references/agent_development.md +++ b/references/agent_development.md @@ -8,11 +8,12 @@ Model facts and protocol rules are linked from [models.md](models.md) and [api_reference.md](api_reference.md). The design advice and prompt examples below are **`application-policy`**: starting points to evaluate on your tasks. -**Contract update checked 2026-10-02:** the 2026-09-18 model change made -`ecnu-max` text-only again; use `ecnu-plus` for vision. Plus now supports -`low` / `medium` / `xhigh` effort when thinking is enabled. The older model -choices below describe the September 12 checks, not current execution advice; -follow [current model selection](models.md) and [thinking](examples.md#thinking-and-tool-history). +**Contract update checked 2026-10-19:** the 2026-10-01 model update moved +`ecnu-max` to a multimodal backend, and both primary models now document image +understanding. Plus supports `low` / `medium` / `xhigh` effort when thinking is +enabled. The older model choices below describe the September 12 checks, not +current execution advice; follow [current model selection](models.md) and +[thinking](examples.md#thinking-and-tool-history). ## Choose a model and reasoning budget @@ -33,9 +34,9 @@ Two upstream differences matter when adapting examples: ECNU still documents `low` / `high` / `max` for direct Chat requests. Use the ECNU protocol mappings; do not invent numeric equivalents. - **Qwen3.8:** upstream thinking defaults and template options differ from - ECNU. On ECNU, thinking defaults off, and `ecnu-plus` ignores effort tiers. - Use `thinking.type`; do not copy `enable_thinking`, `preserve_thinking`, or - `chat_template_kwargs` into ECNU requests. + ECNU. On ECNU, thinking defaults off, and `ecnu-plus` uses `low` / `medium` / + `xhigh` effort tiers. Use `thinking.type`; do not copy `enable_thinking`, + `preserve_thinking`, or `chat_template_kwargs` into ECNU requests. Use the API's thinking control instead of trying to activate it with a "think step by step" phrase. Ask for a concise answer with evidence and checks; diff --git a/references/api_reference.md b/references/api_reference.md index 3ff7ed3..e299038 100644 --- a/references/api_reference.md +++ b/references/api_reference.md @@ -32,6 +32,7 @@ in environment variables; ticket URLs are also credentials. | Anthropic-compatible client | [Anthropic API](https://developer.ecnu.edu.cn/vitepress/llm/api/anthropic.html) | [SDK setup](examples.md#anthropic-sdk); investigate suffix errors only when they occur | | Embed an iframe | [Embed integration](https://developer.ecnu.edu.cn/vitepress/llm/api/embediframe.html) | Do not log or reuse one-time tickets | | Browser link that submits a question | [URL chat](https://developer.ecnu.edu.cn/vitepress/llm/api/urlchat.html) | Opening a submit link sends data; do not auto-open it as a documentation check | +| Error status or response shape | [Error codes](https://developer.ecnu.edu.cn/vitepress/llm/error.html) | [Targeted diagnosis](workflows.md#start-from-the-symptom); keep the actual status and body type | ## Account and changing values diff --git a/references/known_deviations.md b/references/known_deviations.md index 7fe01d0..f33379e 100644 --- a/references/known_deviations.md +++ b/references/known_deviations.md @@ -204,10 +204,11 @@ and credentials were not retained. ## Direct `ecnu-max` image input -Current-contract note (documentation checked 2026-10-02, not a retest): the -[2026-09-18 model change](https://developer.ecnu.edu.cn/vitepress/llm/model.html) -made `ecnu-max` text-only again. Use `ecnu-plus` for image understanding. -The resolved status below describes only the September 12 fixture. +Current-contract note (documentation checked 2026-10-19, not a retest): the +[2026-10-01 model update](https://developer.ecnu.edu.cn/vitepress/llm/model.html) +moved `ecnu-max` to DeepSeek-V4.1-Flash-1017 with native image understanding. +The 2026-09-18 text-only interval is historical, and the resolved status below +still describes only the September 12 fixture. - **Tested at:** 2026-09-12; previous failure on 2026-08-23 - **Environment:** live-2026-09-12-a, direct HTTP; historical live-2026-08-23-a @@ -405,10 +406,12 @@ max default. Its original test date and status are retained. ## Verified coverage on 2026-09-12 through 2026-09-14 -**Current-contract note (2026-10-02):** This is historical coverage, not a -current capability test. The model page now specifies max as text-only with -512K context; use plus for image understanding and the current per-model -thinking tiers. No live revalidation was performed for this note. +**Current-contract note (2026-10-19):** This is historical coverage, not a +current capability test. The model page now documents image understanding on +both primary models; the max-vision result below predates the 2026-09-18 +text-only interval and the 2026-10-01 multimodal backend update. Use the +current per-model thinking tiers. No live revalidation was performed for this +note. Environment `live-2026-09-12-c` made characterization requests across two personal tokens on Windows x64, Node.js 24.14.1, over Asia/Shanghai dates diff --git a/references/models.md b/references/models.md index 400c043..eb82924 100644 --- a/references/models.md +++ b/references/models.md @@ -20,8 +20,12 @@ a measured quality or latency ranking. ## Select the capability, not just a new model name The [2026-09-30 release](https://developer.ecnu.edu.cn/vitepress/llm/release.html) -added multimodal retrieval and image editing. Contract checked 2026-10-02; -these are documented capabilities, not new live-test results. +added multimodal retrieval and image editing. The +[model page](https://developer.ecnu.edu.cn/vitepress/llm/model.html) also +records the 2026-10-01 backend update: `ecnu-max` is now +DeepSeek-V4.1-Flash-1017 and `ecnu-plus` is Qwen3.8-Flash-Next. Contract +checked 2026-10-19; these are documented capabilities, not new live-test +results. | Model | Backend in the current model page | Integration choice | |---|---|---| @@ -31,6 +35,13 @@ these are documented capabilities, not new live-test results. | `ecnu-rerank-vl` | Qwen3-VL-Reranker-8B | Text/image query and candidates; rerank an already retrieved candidate set | | `ecnu-image` | qwen-image-2.1 | Generate from text or edit one existing image; different request formats and prompt handling | +The same page currently lists `ecnu-max` with 512K context, up to 256K output, +and `low` / `high` / `max` effort, and `ecnu-plus` with 256K context, up to +128K output, and `low` / `medium` / `xhigh` effort. Each primary model has a +documented concurrency cap of 3. Recheck those values before displaying or +depending on them. `ecnu-tts` is Fun-CosyVoice3-0.5B with 16 voices and a +4096-character input limit. + Use the [retrieval and image recipes](examples.md) for the differences that affect requests. Equal vector dimensions do not imply compatible embedding spaces: changing embedding models requires re-embedding the indexed items and @@ -38,12 +49,12 @@ using that same model and dimensions for queries. A reranker can change without replacing stored vectors; its relevance scores are not scientific or educational quality judgments and should not be compared across models. -For image understanding, the current model page specifies `ecnu-plus`. -`ecnu-max` currently uses DeepSeek-V4-Flash-0731 and is text-only; older -DeepSeek-V4.1 and successful vision observations are historical, not current -capability guarantees. Keep stable ECNU aliases in requests and recheck backend -labels before displaying them. Do not relabel old generated assets as output -from a newly announced backend. +The current model page documents image understanding on both `ecnu-max` and +`ecnu-plus`. The 2026-10-01 backend update moved `ecnu-max` to +DeepSeek-V4.1-Flash-1017, ending the 2026-09-18 interval when it was text-only. +Keep stable ECNU aliases in requests and recheck backend labels before +displaying them. Do not relabel old generated assets as output from a newly +announced backend. ## ECNU-specific boundaries diff --git a/references/workbuddy_setup.md b/references/workbuddy_setup.md index ec5303d..0aab4a1 100644 --- a/references/workbuddy_setup.md +++ b/references/workbuddy_setup.md @@ -4,7 +4,7 @@ Connect ECNU to WorkBuddy desktop as a custom OpenAI-compatible model source. The desktop implementation and startup behavior below were inspected on 2026-09-12; they have not been revalidated on current WorkBuddy builds. The configuration follows the [ECNU model contract](https://developer.ecnu.edu.cn/vitepress/llm/model.html) -checked on 2026-10-02, but this revised configuration has not been live-tested. +checked on 2026-10-19, but this revised configuration has not been live-tested. ## Configuration file @@ -36,9 +36,12 @@ Edit the JSON file directly. ## Configuration example Use `ecnu-reasoner` for thinking and `ecnu-max` for ordinary text requests. -Both are currently text-only. The input/output limits below are conservative -client budgets: together they stay below the documented 512K context, without -assuming whether K means 1000 or 1024. They are not backend maximums. +The entries below keep this client text-only; since 2026-10-01 the ECNU model +page documents image understanding on both `ecnu-max` and `ecnu-plus`, so +`supportsImages: false` is a client configuration choice, not a model limit. +The input/output limits below are conservative client budgets: together they +stay below the documented 512K context, without assuming whether K means 1000 +or 1024. They are not backend maximums. ```json [ diff --git a/references/workflows.md b/references/workflows.md index 1d82068..a3551ca 100644 --- a/references/workflows.md +++ b/references/workflows.md @@ -26,9 +26,9 @@ These links are dated observations, not claims that the issue still reproduces. | An effort tier fails | Use the chosen model's [documented tiers](examples.md#thinking-and-tool-history); the [September max observation](known_deviations.md#unavailable-reasoning-effort-tiers) does not apply to plus | | TTS error parsing crashes | Tolerate non-JSON errors: [invalid voice](known_deviations.md#invalid-tts-voice-error-shape) | | PCM bytes arrive without format metadata | Configure the format explicitly: [missing headers](known_deviations.md#successful-tts-response-headers) | -| Old max-vision success conflicts with current docs | Current max is text-only; the [September 12 fixture](known_deviations.md#direct-ecnu-max-image-input) predates the September 18 contract change | +| Old max-vision notes conflict with current docs | Current docs list image understanding on both models; the [September 12 fixture](known_deviations.md#direct-ecnu-max-image-input) predates the 2026-09-18 text-only interval and the 2026-10-01 backend update | | `422` | Inspect `detail` and the relevant [request contract](api_reference.md); do not retry the unchanged request | -| `429` | Check [current credits/quota](models.md); stop parallel retries | +| `429` | Check [current credits/quota](models.md) and the documented per-model concurrency cap (3); stop parallel retries | | Timeout or dropped connection after POST | Completion and debit may be unknown; do not automatically resubmit | A later release announcement does not by itself resolve a deviation. Do not From 5c8e6d84947da8e1cf45e7d0127a2642f7c65ae2 Mon Sep 17 00:00:00 2001 From: "J.Jason" <130959319+JJasonSun@users.noreply.github.com> Date: Wed, 21 Oct 2026 16:41:02 +0800 Subject: [PATCH 2/2] docs: correct per-model concurrency limits --- references/models.md | 8 ++++---- references/workflows.md | 2 +- 2 files changed, 5 insertions(+), 5 deletions(-) diff --git a/references/models.md b/references/models.md index eb82924..aad60c7 100644 --- a/references/models.md +++ b/references/models.md @@ -37,10 +37,10 @@ results. The same page currently lists `ecnu-max` with 512K context, up to 256K output, and `low` / `high` / `max` effort, and `ecnu-plus` with 256K context, up to -128K output, and `low` / `medium` / `xhigh` effort. Each primary model has a -documented concurrency cap of 3. Recheck those values before displaying or -depending on them. `ecnu-tts` is Fun-CosyVoice3-0.5B with 16 voices and a -4096-character input limit. +128K output, and `low` / `medium` / `xhigh` effort. Per-user concurrency caps +are 3 for `ecnu-max` and 5 for `ecnu-plus`, counted independently. Recheck +those values before displaying or depending on them. `ecnu-tts` is +Fun-CosyVoice3-0.5B with 16 voices and a 4096-character input limit. Use the [retrieval and image recipes](examples.md) for the differences that affect requests. Equal vector dimensions do not imply compatible embedding diff --git a/references/workflows.md b/references/workflows.md index a3551ca..adc82d3 100644 --- a/references/workflows.md +++ b/references/workflows.md @@ -28,7 +28,7 @@ These links are dated observations, not claims that the issue still reproduces. | PCM bytes arrive without format metadata | Configure the format explicitly: [missing headers](known_deviations.md#successful-tts-response-headers) | | Old max-vision notes conflict with current docs | Current docs list image understanding on both models; the [September 12 fixture](known_deviations.md#direct-ecnu-max-image-input) predates the 2026-09-18 text-only interval and the 2026-10-01 backend update | | `422` | Inspect `detail` and the relevant [request contract](api_reference.md); do not retry the unchanged request | -| `429` | Check [current credits/quota](models.md) and the documented per-model concurrency cap (3); stop parallel retries | +| `429` | Check [current credits/quota](models.md) and the independent per-user concurrency caps (`ecnu-max`: 3; `ecnu-plus`: 5); stop parallel retries | | Timeout or dropped connection after POST | Completion and debit may be unknown; do not automatically resubmit | A later release announcement does not by itself resolve a deviation. Do not