Skip to content

Commit 403cf31

Browse files
committed
fix(providers): curate Volcengine model catalogs
1 parent 8087691 commit 403cf31

7 files changed

Lines changed: 138 additions & 29 deletions

File tree

‎docs-site/src/content/docs/guides/providers.md‎

Lines changed: 6 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -224,6 +224,12 @@ Volcengine Agent Plan uses its native Responses endpoint through `openai-respons
224224
> `volcengine-coding-plan` consumes Coding Plan quota, and `volcengine-agent-plan` consumes Agent
225225
> Plan quota. Use the key and endpoint issued for the same product; the ordinary `/api/v3` endpoint
226226
> can incur pay-as-you-go charges even when a Plan subscription exists.
227+
> The presets use curated static model catalogs because Ark's `/models` response also includes
228+
> embedding, image, video, and 3D resources, the Coding gateway returns that same broad catalog,
229+
> and the Agent Plan gateway has no `/models` resource. Pay-as-you-go defaults to
230+
> `doubao-seed-2-1-pro-260628`; its curated catalog also includes current DeepSeek and GLM text
231+
> models. Coding Plan defaults to `ark-code-latest`, while Agent Plan defaults to
232+
> `deepseek-v4-pro`.
227233
228234
> **Tencent Cloud Coding Plan usage restriction:** Tencent documents this subscription for
229235
> interactive coding tools only. General API automation, custom application backends, and

‎docs-site/src/content/docs/ja/guides/providers.md‎

Lines changed: 6 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -158,6 +158,12 @@ Volcengine Agent Plan は `openai-responses` アダプターでネイティブ R
158158
> `volcengine-coding-plan` は Coding Plan の割り当て、`volcengine-agent-plan` は Agent Plan
159159
> の割り当てを使用します。同じ製品で発行されたキーとエンドポイントを組み合わせてください。
160160
> Plan 契約があっても通常の `/api/v3` 呼び出しには従量課金が発生する場合があります。
161+
> 3 つの preset は選別済みの静的モデルカタログを使用します。Ark の `/models` はテキストに加えて
162+
> Embedding、画像、動画、3D リソースも返し、Coding ゲートウェイも同じ広範なカタログを返します。
163+
> Agent Plan ゲートウェイには `/models` リソースがありません。従量課金のデフォルトは
164+
> `doubao-seed-2-1-pro-260628` で、静的カタログには現在の DeepSeek と GLM のテキストモデルも
165+
> 含まれます。Coding Plan のデフォルトは `ark-code-latest`、Agent Plan は
166+
> `deepseek-v4-pro` です。
161167
162168
> **Tencent Cloud Coding Plan の利用制限:** Tencent はこのサブスクリプションを対話型
163169
> コーディングツール専用としています。一般的な API 自動化、カスタムアプリのバックエンド、

‎docs-site/src/content/docs/ko/guides/providers.md‎

Lines changed: 5 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -158,6 +158,11 @@ Volcengine Agent Plan은 `openai-responses` 어댑터로 네이티브 Responses
158158
> `volcengine-coding-plan`은 Coding Plan 할당량, `volcengine-agent-plan`은 Agent Plan
159159
> 할당량을 사용합니다. 같은 상품에서 발급된 키와 엔드포인트를 함께 사용해야 하며, Plan 구독이
160160
> 있어도 일반 `/api/v3` 엔드포인트 호출에는 종량제 요금이 발생할 수 있습니다.
161+
> 세 preset은 선별된 정적 모델 카탈로그를 사용합니다. Ark의 `/models`는 텍스트와 함께
162+
> Embedding, 이미지, 비디오, 3D 리소스도 반환하고 Coding 게이트웨이도 같은 광범위한 카탈로그를
163+
> 반환합니다. Agent Plan 게이트웨이에는 `/models` 리소스가 없습니다. 종량제 기본값은
164+
> `doubao-seed-2-1-pro-260628`이며 정적 카탈로그에는 현재 DeepSeek와 GLM 텍스트 모델도
165+
> 포함됩니다. Coding Plan의 기본값은 `ark-code-latest`, Agent Plan은 `deepseek-v4-pro`입니다.
161166
162167
> **Tencent Cloud Coding Plan 사용 제한:** Tencent는 이 구독을 대화형 코딩 도구 전용으로
163168
> 안내합니다. 일반 API 자동화, 사용자 애플리케이션 백엔드 및 비대화형 일괄 호출은 금지되며

‎docs-site/src/content/docs/ru/guides/providers.md‎

Lines changed: 6 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -169,6 +169,12 @@ Volcengine Agent Plan использует нативную конечную т
169169
> `volcengine-coding-plan` расходует квоту Coding Plan, а `volcengine-agent-plan` — квоту Agent
170170
> Plan. Используйте ключ и конечную точку одного продукта; обычный `/api/v3` может тарифицироваться
171171
> отдельно даже при активной подписке Plan.
172+
> Все три preset используют отобранные статические каталоги моделей. Ark `/models` возвращает
173+
> текстовые, embedding-, графические, видео- и 3D-ресурсы, а шлюз Coding выдаёт тот же широкий
174+
> каталог. У шлюза Agent Plan ресурса `/models` нет. Для pay-as-you-go модель по умолчанию —
175+
> `doubao-seed-2-1-pro-260628`; его статический каталог также включает актуальные текстовые модели
176+
> DeepSeek и GLM. Для Coding Plan модель по умолчанию — `ark-code-latest`, для Agent Plan —
177+
> `deepseek-v4-pro`.
172178
173179
> **Ограничение Tencent Cloud Coding Plan:** Tencent разрешает использовать эту подписку только
174180
> в интерактивных инструментах программирования. Автоматизация общего API, серверы пользовательских

‎docs-site/src/content/docs/zh-cn/guides/providers.md‎

Lines changed: 5 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -148,6 +148,11 @@ ChatGPT 转发预设。仪表盘的 **Add provider** 选择器会打开密钥提
148148
> **三条火山方舟计费线路:**`volcengine` 是按量付费方舟 API,`volcengine-coding-plan`
149149
> 消耗 Coding Plan 额度,`volcengine-agent-plan` 消耗 Agent Plan 额度。密钥与端点需要属于
150150
> 同一产品;已经订阅 Plan 时调用普通 `/api/v3` 端点仍可能产生按量费用。
151+
> 三个 preset 使用经过筛选的静态模型目录:方舟 `/models` 同时返回文本、Embedding、图片、
152+
> 视频和 3D 资源,Coding 网关也会返回这份宽泛目录,Agent Plan 网关没有 `/models` 资源。
153+
> 按量付费默认使用 `doubao-seed-2-1-pro-260628`,静态目录还包含当前 DeepSeek 和 GLM
154+
> 文本模型。Coding Plan 默认使用 `ark-code-latest`,Agent Plan 默认使用
155+
> `deepseek-v4-pro`。
151156
152157
> **腾讯云 Coding Plan 使用限制:**腾讯将此订阅限定为交互式编程工具使用。禁止通用 API
153158
> 自动化、自定义应用后端和非交互式批量调用;违规使用可能导致套餐密钥被停用。

‎src/providers/registry.ts‎

Lines changed: 66 additions & 16 deletions
Original file line numberDiff line numberDiff line change
@@ -255,6 +255,41 @@ const ALIBABA_INTL_TOKEN_PLAN_QWEN_MODELS = [
255255
// coding tools (not custom application backends or non-interactive batch automation).
256256
// Evidence: https://cloud.tencent.cn/document/product/1823/130092
257257
const TENCENT_CODING_PLAN_MODELS = ["tc-code-latest", "glm-5", "kimi-k2.5", "minimax-m2.5"];
258+
// Volcengine's authenticated /api/v3/models catalog mixes chat models with embedding,
259+
// image, video, and 3D generation resources. Keep the Codex-facing presets scoped to
260+
// models documented for text/agent or Coding Plan use.
261+
const VOLCENGINE_ARK_MODELS = [
262+
"doubao-seed-2-1-pro-260628",
263+
"doubao-seed-2-1-turbo-260628",
264+
"doubao-seed-evolving",
265+
"deepseek-v4-pro-260425",
266+
"deepseek-v4-flash-260425",
267+
"deepseek-v3-2-251201",
268+
"glm-5-2-260617",
269+
"glm-4-7-251222",
270+
];
271+
const VOLCENGINE_DOUBAO_THINKING_MODELS = [
272+
"doubao-seed-2-1-pro-260628",
273+
"doubao-seed-2-1-turbo-260628",
274+
"doubao-seed-evolving",
275+
];
276+
const VOLCENGINE_CODING_PLAN_MODELS = [
277+
"ark-code-latest",
278+
"doubao-seed-2.0-code",
279+
"deepseek-v4-pro",
280+
"deepseek-v4-flash",
281+
"glm-5.1",
282+
"kimi-k2.6",
283+
"minimax-m3",
284+
];
285+
const VOLCENGINE_AGENT_PLAN_MODELS = [
286+
"deepseek-v4-pro",
287+
"deepseek-v4-flash",
288+
"glm-5.1",
289+
"kimi-k2.6",
290+
"minimax-m3",
291+
"doubao-seed-2.0-pro",
292+
];
258293
const ALIBABA_INTL_TOKEN_PLAN_INPUT_MODALITIES: Record<string, string[]> = {
259294
"qwen3.8-max-preview": ["text", "image"],
260295
"qwen3.7-max": ["text", "image"],
@@ -896,17 +931,30 @@ export const PROVIDER_REGISTRY: readonly ProviderRegistryEntry[] = [
896931
adapter: "openai-chat",
897932
authKind: "key",
898933
dashboardUrl: "https://console.volcengine.com/ark/region:ark+cn-beijing/apikey",
899-
defaultModel: "doubao-seed-2-0-lite-260215",
900-
models: ["doubao-seed-2-0-lite-260215"],
901-
liveModels: true,
902-
modelReasoningEfforts: {
903-
"doubao-seed-2-0-lite-260215": THINKING_TOGGLE_EFFORTS,
904-
},
905-
modelReasoningEffortMap: {
906-
"doubao-seed-2-0-lite-260215": THINKING_TOGGLE_MAP,
907-
},
908-
thinkingToggleModels: ["doubao-seed-2-0-lite-260215"],
909-
note: "Pay-as-you-go Ark API. Calls on this endpoint do not consume Coding Plan or Agent Plan quota.",
934+
defaultModel: "doubao-seed-2-1-pro-260628",
935+
models: VOLCENGINE_ARK_MODELS,
936+
liveModels: false,
937+
modelReasoningEfforts: Object.fromEntries(
938+
VOLCENGINE_DOUBAO_THINKING_MODELS.map(id => [id, THINKING_TOGGLE_EFFORTS]),
939+
),
940+
modelReasoningEffortMap: Object.fromEntries(
941+
VOLCENGINE_DOUBAO_THINKING_MODELS.map(id => [id, THINKING_TOGGLE_MAP]),
942+
),
943+
thinkingToggleModels: VOLCENGINE_DOUBAO_THINKING_MODELS,
944+
preserveReasoningContentModels: [
945+
"deepseek-v4-pro-260425",
946+
"deepseek-v4-flash-260425",
947+
"glm-5-2-260617",
948+
"glm-4-7-251222",
949+
],
950+
noVisionModels: [
951+
"deepseek-v4-pro-260425",
952+
"deepseek-v4-flash-260425",
953+
"deepseek-v3-2-251201",
954+
"glm-5-2-260617",
955+
"glm-4-7-251222",
956+
],
957+
note: "Pay-as-you-go Ark API with a curated text/agent catalog. Calls on this endpoint do not consume Coding Plan or Agent Plan quota.",
910958
},
911959
{
912960
id: "volcengine-coding-plan",
@@ -916,9 +964,9 @@ export const PROVIDER_REGISTRY: readonly ProviderRegistryEntry[] = [
916964
authKind: "key",
917965
dashboardUrl: "https://console.volcengine.com/ark/region:ark+cn-beijing/overview",
918966
defaultModel: "ark-code-latest",
919-
models: ["ark-code-latest"],
920-
liveModels: true,
921-
note: "Coding Plan subscription endpoint. Use the plan key issued by the Ark console.",
967+
models: VOLCENGINE_CODING_PLAN_MODELS,
968+
liveModels: false,
969+
note: "Coding Plan subscription endpoint with plan-scoped model aliases. Use the plan key issued by the Ark console.",
922970
},
923971
{
924972
id: "volcengine-agent-plan",
@@ -928,8 +976,10 @@ export const PROVIDER_REGISTRY: readonly ProviderRegistryEntry[] = [
928976
adapter: "openai-responses",
929977
authKind: "key",
930978
dashboardUrl: "https://console.volcengine.com/ark/region:ark+cn-beijing/overview",
931-
liveModels: true,
932-
note: "Agent Plan subscription endpoint over the native Responses API.",
979+
defaultModel: "deepseek-v4-pro",
980+
models: VOLCENGINE_AGENT_PLAN_MODELS,
981+
liveModels: false,
982+
note: "Agent Plan subscription endpoint over the native Responses API with a static fallback catalog.",
933983
},
934984
// 2026-07-10: docs unverified; model data frozen. Evidence: devlog/_plan/260710_provider_hardening/002_research_cn.md.
935985
{ id: "qianfan", label: "Qianfan (Baidu)", baseUrl: "https://qianfan.baidubce.com/v2", adapter: "openai-chat", authKind: "key", dashboardUrl: "https://console.bce.baidu.com/iam/#/iam/apikey/list" },

‎tests/volcengine-providers.test.ts‎

Lines changed: 44 additions & 13 deletions
Original file line numberDiff line numberDiff line change
@@ -15,46 +15,77 @@ describe("Volcengine Ark providers", () => {
1515
baseUrl: "https://ark.cn-beijing.volces.com/api/v3",
1616
adapter: "openai-chat",
1717
authKind: "key",
18-
defaultModel: "doubao-seed-2-0-lite-260215",
19-
models: ["doubao-seed-2-0-lite-260215"],
20-
liveModels: true,
21-
thinkingToggleModels: ["doubao-seed-2-0-lite-260215"],
18+
defaultModel: "doubao-seed-2-1-pro-260628",
19+
models: [
20+
"doubao-seed-2-1-pro-260628",
21+
"doubao-seed-2-1-turbo-260628",
22+
"doubao-seed-evolving",
23+
"deepseek-v4-pro-260425",
24+
"deepseek-v4-flash-260425",
25+
"deepseek-v3-2-251201",
26+
"glm-5-2-260617",
27+
"glm-4-7-251222",
28+
],
29+
liveModels: false,
30+
thinkingToggleModels: [
31+
"doubao-seed-2-1-pro-260628",
32+
"doubao-seed-2-1-turbo-260628",
33+
"doubao-seed-evolving",
34+
],
2235
});
2336
expect(PROVIDER_REGISTRY.find(provider => provider.id === "volcengine-coding-plan")).toMatchObject({
2437
label: "Volcengine Ark Coding Plan",
2538
baseUrl: "https://ark.cn-beijing.volces.com/api/coding/v3",
2639
adapter: "openai-chat",
2740
authKind: "key",
2841
defaultModel: "ark-code-latest",
29-
models: ["ark-code-latest"],
30-
liveModels: true,
42+
models: [
43+
"ark-code-latest",
44+
"doubao-seed-2.0-code",
45+
"deepseek-v4-pro",
46+
"deepseek-v4-flash",
47+
"glm-5.1",
48+
"kimi-k2.6",
49+
"minimax-m3",
50+
],
51+
liveModels: false,
3152
});
3253
expect(PROVIDER_REGISTRY.find(provider => provider.id === "volcengine-agent-plan")).toMatchObject({
3354
label: "Volcengine Ark Agent Plan",
3455
baseUrl: "https://ark.cn-beijing.volces.com/api/plan/v3",
3556
responsesPath: "/responses",
3657
adapter: "openai-responses",
3758
authKind: "key",
38-
liveModels: true,
59+
defaultModel: "deepseek-v4-pro",
60+
models: [
61+
"deepseek-v4-pro",
62+
"deepseek-v4-flash",
63+
"glm-5.1",
64+
"kimi-k2.6",
65+
"minimax-m3",
66+
"doubao-seed-2.0-pro",
67+
],
68+
liveModels: false,
3969
});
4070
});
4171

4272
test("derives key login and dashboard presets from the canonical registry", () => {
4373
expect(KEY_LOGIN_PROVIDERS.volcengine).toMatchObject({
4474
baseUrl: "https://ark.cn-beijing.volces.com/api/v3",
45-
defaultModel: "doubao-seed-2-0-lite-260215",
46-
liveModels: true,
75+
defaultModel: "doubao-seed-2-1-pro-260628",
76+
liveModels: false,
4777
});
4878
expect(KEY_LOGIN_PROVIDERS["volcengine-coding-plan"]).toMatchObject({
4979
baseUrl: "https://ark.cn-beijing.volces.com/api/coding/v3",
5080
defaultModel: "ark-code-latest",
51-
liveModels: true,
81+
liveModels: false,
5282
});
5383
expect(KEY_LOGIN_PROVIDERS["volcengine-agent-plan"]).toMatchObject({
5484
baseUrl: "https://ark.cn-beijing.volces.com/api/plan/v3",
5585
responsesPath: "/responses",
5686
adapter: "openai-responses",
57-
liveModels: true,
87+
defaultModel: "deepseek-v4-pro",
88+
liveModels: false,
5889
});
5990
for (const id of ["volcengine", "volcengine-coding-plan", "volcengine-agent-plan"]) {
6091
expect(deriveProviderPresets().find(provider => provider.id === id)).toMatchObject({ auth: "key" });
@@ -74,7 +105,7 @@ describe("Volcengine Ark providers", () => {
74105
},
75106
},
76107
};
77-
const route = routeModel(config, "volcengine-agent-plan/ark-agent-latest");
108+
const route = routeModel(config, "volcengine-agent-plan/deepseek-v4-pro");
78109
expect(route.provider.responsesPath).toBe("/responses");
79110

80111
const request = createResponsesPassthroughAdapter(route.provider).buildRequest({
@@ -100,7 +131,7 @@ describe("Volcengine Ark providers", () => {
100131
},
101132
},
102133
};
103-
const route = routeModel(config, "volcengine/doubao-seed-2-0-lite-260215");
134+
const route = routeModel(config, "volcengine/doubao-seed-2-1-pro-260628");
104135
const request = createOpenAIChatAdapter(route.provider).buildRequest({
105136
modelId: route.modelId,
106137
context: {

0 commit comments

Comments
 (0)