A local-first multi-agent app development platform, built on PenguinHarness
feat(core): TokenDance's free Dots3-Note Preview joins the catalog (#730)
Yaowei Zhengfff691fb6afe
diff --git a/changelog/unreleased/2026-09-15-tokendance-dots-3-note-preview.md b/changelog/unreleased/2026-09-15-tokendance-dots-3-note-preview.mdnew file mode 100644index 00000000..0d251ed5--- /dev/null+++ b/changelog/unreleased/2026-09-15-tokendance-dots-3-note-preview.md@@ -0,0 +1,11 @@+# TokenDance's free Dots3-Note Preview joins the catalog++- **Date:** 2026-09-15+- **Type:** feat+- **Scope:** `core`++[中文版](2026-09-15-tokendance-dots-3-note-preview.zh.md)++## What changed++- The TokenDance group gains `dots-3-note-preview`, Dots Studio's Dots3-Note Preview: free on the gateway (CNY 0 on every bucket, so costs compute to 0 and the free badge shows), a 512K context window, and the vision flag set from the gateway's own listing, which describes the model as covering multimodal understanding. Existing Projects pick it up through the models page's "sync presets".diff --git a/changelog/unreleased/2026-09-15-tokendance-dots-3-note-preview.zh.md b/changelog/unreleased/2026-09-15-tokendance-dots-3-note-preview.zh.mdnew file mode 100644index 00000000..87af5b3e--- /dev/null+++ b/changelog/unreleased/2026-09-15-tokendance-dots-3-note-preview.zh.md@@ -0,0 +1,11 @@+# TokenDance 的免费模型 Dots3-Note Preview 进入目录++- **Date:** 2026-09-15+- **Type:** feat+- **Scope:** `core`++[English](2026-09-15-tokendance-dots-3-note-preview.md)++## 变更内容++- TokenDance 分组新增 `dots-3-note-preview`,即 Dots Studio 的 Dots3-Note Preview:网关免费(三桶价格均为 0 元,成本按 0 计、卡片显示免费徽标),512K 上下文,视觉标志按网关自己的目录描述(「多模态理解」)设为支持。已有 Project 经模型库的「同步预置」取得它。diff --git a/packages/core/src/state/model-catalog.ts b/packages/core/src/state/model-catalog.tsindex b0e42362..fbeaa0b6 100644--- a/packages/core/src/state/model-catalog.ts+++ b/packages/core/src/state/model-catalog.ts@@ -1336,6 +1336,27 @@ export const MODEL_CATALOG: ModelCatalogEntry[] = [clientType: "openai-chat",baseUrl: TOKENDANCE_BASE_URL,},+ {+ // Dots Studio's (rednote / Xiaohongshu) Dots3-Note Preview: an open-weight MoE, 280B total+ // and 16B active, the lightest of the Dots 3 family. TokenDance lists it as "Dots3-Note+ // Preview(Free)", and the name is kept as the seller spells it, tag and full-width+ // parentheses included, as the OpenRouter `(free)` rows keep theirs. The price is a+ // genuine CNY 0 on every bucket, the same treatment as the OpenRouter `:free` rows, so+ // costs compute to 0 and the free badge shows — over a 512,000-token context window, with+ // openai:chat-completions and anthropic:messages as its supported_protocols (this group's+ // openai-chat pin is a convention here, not the only shape the id serves). The gateway's+ // own listing describes it as covering multimodal understanding, which is what the+ // vision flag records; no image request was sent to it. Read 2026-09-15 from the+ // gateway's /models listing and tokendance.space/models/dots-3-note-preview.+ modelId: "dots-3-note-preview",+ displayName: "Dots3-Note Preview(Free)",+ provider: "tokendance",+ contextWindow: 512000,+ pricing: cny(0, 0, 0),+ supportsVision: true,+ clientType: "openai-chat",+ baseUrl: TOKENDANCE_BASE_URL,+ },{modelId: "glm-5.3",displayName: "GLM-5.3",diff --git a/packages/core/test/model-catalog.test.ts b/packages/core/test/model-catalog.test.tsindex 8a73a180..143336a6 100644--- a/packages/core/test/model-catalog.test.ts+++ b/packages/core/test/model-catalog.test.ts@@ -119,11 +119,17 @@ describe("model-catalog", () => {it("every entry has valid three-bucket pricing; context_window is a positive integer", () => {for (const m of MODEL_CATALOG) {- if (m.provider === "vllm" || m.modelId.endsWith(":free") || m.modelId === "openrouter/free") {+ if (+ m.provider === "vllm" ||+ m.modelId.endsWith(":free") ||+ m.modelId === "openrouter/free" ||+ m.modelId === "dots-3-note-preview"+ ) {// Self-hosted vLLM and the free-tier gateway rows share one treatment: a genuine $0// price (not "unknown"), so costs compute to 0 and the free badge shows. Nobody bills// per token for either — a vLLM deployment costs its operator hardware, which no- // catalog rate expresses.+ // catalog rate expresses. TokenDance's dots-3-note-preview is the one free row of+ // its group.expect(m.pricing, m.modelId).toBeDefined();expect([m.pricing!.cache_read, m.pricing!.cache_write, m.pricing!.output]).toEqual([0, 0, 0,@@ -421,6 +427,7 @@ describe("model-catalog", () => {["deepseek-v4-flash-vision-exp", 1000000, true],["deepseek-v4-pro-0813", 1000000, false],["deepseek-v4.1-flash", 1000000, true],+ ["dots-3-note-preview", 512000, true],["glm-5.3", 1000000, false],["glm-5.3-flash", 1000000, true],["hy4-preview", 1024000, false],@@ -497,6 +504,17 @@ describe("model-catalog", () => {0.005714, 0.285714, 1.142857,]);}+ // The one free row of the group: CNY 0 on every bucket, no discount decoration, a 512K+ // window, and the seller's own spelling of the name, its "(Free)" tag included, as the+ // OpenRouter "(free)" rows keep theirs.+ const dots = td.find((m) => m.modelId === "dots-3-note-preview")!;+ expect(dots.displayName).toBe("Dots3-Note Preview(Free)");+ expect(dots.contextWindow).toBe(512000);+ expect(dots.discount).toBeUndefined();+ expect([dots.pricing!.cache_read, dots.pricing!.cache_write, dots.pricing!.output]).toEqual([+ 0, 0, 0,+ ]);+ expect(dots.supportsVision).toBe(true);// Display names are the seller's own spelling, not a prettified one.expect(td.filter((m) => m.modelId.startsWith("seed-")).map((m) => m.displayName)).toEqual(["Seed-2.1-Pro",diff --git a/packages/docs/content/models.en.md b/packages/docs/content/models.en.mdindex 97994420..dcf556d4 100644--- a/packages/docs/content/models.en.md+++ b/packages/docs/content/models.en.md@@ -80,7 +80,7 @@ The gateway groups (openrouter / fireworks / siliconflow / tokendance / qwen-payThe preset catalog also carries OpenRouter's free tier: the `:free` model variant `nvidia/nemotron-3-ultra-550b-a55b:free` and the `openrouter/free` unified Free Models Router. They cost nothing, but are subject to OpenRouter's free-tier rate limits and data policy.-Some models in the preset catalog: deepseek-flash / deepseek-v4-pro / deepseek-v4-flash / deepseek-v4-flash-vision-exp (`deepseek-flash` and `deepseek-v4-flash-vision-exp` read images; `deepseek-v4-flash` and `deepseek-v4-pro` are text-only. Since 2026-09-10 the two legacy Flash ids are retired names that DeepSeek serves from V4.1 Flash at the Flash price, which changes what they cost but not what V4 Flash reads: send an image to `deepseek-flash`), MiniMax-M3, gemini-3.8-flash, claude-opus-5 / claude-opus-4-8 / claude-sonnet-5, gpt-6-astra / gpt-5.6 / gpt-5.5, glm-5.3 / glm-5.3-flash, kimi-k3, qwen3.8-max / qwen3.8-flash, seed-2.1-pro / seed-2.1-turbo / seed-evolving (not exhaustive). The whole OpenAI line-up is listed twice — directly (your own OpenAI key, list prices) and on OpenRouter as `openai/<id>` (the gateway's rates, which follow its running promotions). DeepSeek's direct-group rows record the official peak tier and declare its off-peak schedule: outside Beijing weekday 9:00–12:00 and 14:00–18:00 every bucket is halved, which the models page marks with a `50% off` badge and the cost center bills at. Three resold rows follow the same schedule, because their sellers pass DeepSeek's own windows through: TokenDance's `deepseek-v4.1-flash` and `deepseek-v4-flash-vision-exp`, and OpenRouter's `deepseek/deepseek-v4.1-flash`. The stored price is always the peak one, so what is on disk does not depend on the hour a Project was created or re-synced in. `glm-5.3-flash` appears three times, and all three rows accept images: AgentHub's GLM client forwards image parts for this one GLM id (every other GLM id refuses them), while the OpenRouter row `z-ai/glm-5.3-flash` and the TokenDance row go through the generic OpenAI-compatible client, which carries them for any id. What the three rows do not share is the price: each records what its own seller charges, so they disagree while a promotion is running.+Some models in the preset catalog: deepseek-flash / deepseek-v4-pro / deepseek-v4-flash / deepseek-v4-flash-vision-exp (`deepseek-flash` and `deepseek-v4-flash-vision-exp` read images; `deepseek-v4-flash` and `deepseek-v4-pro` are text-only. Since 2026-09-10 the two legacy Flash ids are retired names that DeepSeek serves from V4.1 Flash at the Flash price, which changes what they cost but not what V4 Flash reads: send an image to `deepseek-flash`), MiniMax-M3, gemini-3.8-flash, claude-opus-5 / claude-opus-4-8 / claude-sonnet-5, gpt-6-astra / gpt-5.6 / gpt-5.5, glm-5.3 / glm-5.3-flash, kimi-k3, qwen3.8-max / qwen3.8-flash, seed-2.1-pro / seed-2.1-turbo / seed-evolving, dots-3-note-preview (free on TokenDance, 512K context) (not exhaustive). The whole OpenAI line-up is listed twice — directly (your own OpenAI key, list prices) and on OpenRouter as `openai/<id>` (the gateway's rates, which follow its running promotions). DeepSeek's direct-group rows record the official peak tier and declare its off-peak schedule: outside Beijing weekday 9:00–12:00 and 14:00–18:00 every bucket is halved, which the models page marks with a `50% off` badge and the cost center bills at. Three resold rows follow the same schedule, because their sellers pass DeepSeek's own windows through: TokenDance's `deepseek-v4.1-flash` and `deepseek-v4-flash-vision-exp`, and OpenRouter's `deepseek/deepseek-v4.1-flash`. The stored price is always the peak one, so what is on disk does not depend on the hour a Project was created or re-synced in. `glm-5.3-flash` appears three times, and all three rows accept images: AgentHub's GLM client forwards image parts for this one GLM id (every other GLM id refuses them), while the OpenRouter row `z-ai/glm-5.3-flash` and the TokenDance row go through the generic OpenAI-compatible client, which carries them for any id. What the three rows do not share is the price: each records what its own seller charges, so they disagree while a promotion is running.TokenDance rows record the gateway's list price and, where a promotion is running, the rate off it. Nine models are discounted today — `deepseek-v4-flash-0731`, `deepseek-v4-pro-0813` and `kimi-k3` at 20% off, `glm-5.3`, `glm-5.3-flash` and `qwen3.8-max` at 10%, and the three Doubao Seed rows (`seed-2.1-pro`, `seed-2.1-turbo`, `seed-evolving`) at 50%. Two further rows carry no flat discount and instead follow DeepSeek's own off-peak schedule, described above. Their model cards show the price being billed right now, with the rate as a badge, and a Project is preset with the **discounted** price, so the cost center charges what the gateway charges. Prices you edit yourself keep the discount decoration off the card: the figure is then yours, not the gateway's.diff --git a/packages/docs/content/models.zh.md b/packages/docs/content/models.zh.mdindex eb0f22ef..8a15fc01 100644--- a/packages/docs/content/models.zh.md+++ b/packages/docs/content/models.zh.md@@ -80,7 +80,7 @@ api_key = "sk-..."预置目录还收录了 OpenRouter 的免费档:`:free` 模型变体 `nvidia/nemotron-3-ultra-550b-a55b:free` 与统一路由 `openrouter/free`(Free Models Router),零成本可用,但受 OpenRouter 免费档速率限制与数据政策约束。-预置目录中的部分模型:deepseek-flash / deepseek-v4-pro / deepseek-v4-flash / deepseek-v4-flash-vision-exp(其中 `deepseek-flash`、`deepseek-v4-flash-vision-exp` 支持图像输入,`deepseek-v4-flash` 与 `deepseek-v4-pro` 为纯文本。自 2026-09-10 起,两个 V4 Flash id 已是退役名称,DeepSeek 改由 V4.1 Flash 承接、按 Flash 价计费——变的是价格,不是 V4 Flash 能读什么:要发图请用 `deepseek-flash`)、MiniMax-M3、gemini-3.8-flash、claude-opus-5 / claude-opus-4-8 / claude-sonnet-5、gpt-6-astra / gpt-5.6 / gpt-5.5、glm-5.3 / glm-5.3-flash、kimi-k3、qwen3.8-max / qwen3.8-flash、seed-2.1-pro / seed-2.1-turbo / seed-evolving 等(非完整清单)。OpenAI 全系列都收录了两份——直连(用自己的 OpenAI Key,记牌价)与 OpenRouter 上的 `openai/<id>`(记网关实际计费价,会随其促销浮动)。DeepSeek 直连分组的条目记录官方高峰档,并声明其空闲时段规则:在北京时间周一至周五 9:00–12:00、14:00–18:00 之外,各档价格减半——模型页以「省 50%」徽标标记,成本中心按此计价。另有三条转售条目沿用同一套时段,因为卖家原样传递了 DeepSeek 自己的窗口:TokenDance 的 `deepseek-v4.1-flash` 与 `deepseek-v4-flash-vision-exp`,以及 OpenRouter 的 `deepseek/deepseek-v4.1-flash`。存盘的始终是高峰价,因此磁盘上的数字不会随 Project 创建或同步预置的时刻而变。`glm-5.3-flash` 收录了三份,三条都支持图像输入:AgentHub 的 GLM 客户端只为这一个 GLM id 转发图像部件(其余 GLM id 一律拒绝),而 OpenRouter 上的 `z-ai/glm-5.3-flash` 与 TokenDance 分组的同名条目走通用 OpenAI 兼容客户端,对任何 id 都能携带图片。三条不一致的是价格:每一条都记录各自卖家的收费,因此促销期间三者不同。+预置目录中的部分模型:deepseek-flash / deepseek-v4-pro / deepseek-v4-flash / deepseek-v4-flash-vision-exp(其中 `deepseek-flash`、`deepseek-v4-flash-vision-exp` 支持图像输入,`deepseek-v4-flash` 与 `deepseek-v4-pro` 为纯文本。自 2026-09-10 起,两个 V4 Flash id 已是退役名称,DeepSeek 改由 V4.1 Flash 承接、按 Flash 价计费——变的是价格,不是 V4 Flash 能读什么:要发图请用 `deepseek-flash`)、MiniMax-M3、gemini-3.8-flash、claude-opus-5 / claude-opus-4-8 / claude-sonnet-5、gpt-6-astra / gpt-5.6 / gpt-5.5、glm-5.3 / glm-5.3-flash、kimi-k3、qwen3.8-max / qwen3.8-flash、seed-2.1-pro / seed-2.1-turbo / seed-evolving、dots-3-note-preview(TokenDance 上免费,512K 上下文)等(非完整清单)。OpenAI 全系列都收录了两份——直连(用自己的 OpenAI Key,记牌价)与 OpenRouter 上的 `openai/<id>`(记网关实际计费价,会随其促销浮动)。DeepSeek 直连分组的条目记录官方高峰档,并声明其空闲时段规则:在北京时间周一至周五 9:00–12:00、14:00–18:00 之外,各档价格减半——模型页以「省 50%」徽标标记,成本中心按此计价。另有三条转售条目沿用同一套时段,因为卖家原样传递了 DeepSeek 自己的窗口:TokenDance 的 `deepseek-v4.1-flash` 与 `deepseek-v4-flash-vision-exp`,以及 OpenRouter 的 `deepseek/deepseek-v4.1-flash`。存盘的始终是高峰价,因此磁盘上的数字不会随 Project 创建或同步预置的时刻而变。`glm-5.3-flash` 收录了三份,三条都支持图像输入:AgentHub 的 GLM 客户端只为这一个 GLM id 转发图像部件(其余 GLM id 一律拒绝),而 OpenRouter 上的 `z-ai/glm-5.3-flash` 与 TokenDance 分组的同名条目走通用 OpenAI 兼容客户端,对任何 id 都能携带图片。三条不一致的是价格:每一条都记录各自卖家的收费,因此促销期间三者不同。TokenDance 分组的条目记录该网关的牌价,并在有促销时记录折扣率。当前有九个模型处于折扣中——`deepseek-v4-flash-0731`、`deepseek-v4-pro-0813` 与 `kimi-k3` 八折,`glm-5.3`、`glm-5.3-flash` 与 `qwen3.8-max` 九折,三条 Doubao Seed 条目(`seed-2.1-pro`、`seed-2.1-turbo`、`seed-evolving`)五折。另有两条条目不走固定折扣,而是沿用上文所述 DeepSeek 自己的空闲时段规则。它们的模型卡片显示当前实际计费的那个价格,并以徽标标出折扣率;新建 Project 预置的是**折后价**,因此成本中心按网关实际收费计价。自行修改过价格的条目不再显示折扣标记:此时那个数字属于你,而不是网关。