mirror of
https://github.com/getcompanion-ai/co-mono.git
synced 2026-04-15 22:03:45 +00:00
Migrate zai provider from Anthropic to OpenAI-compatible API
- Migrate glm-4.5, glm-4.5-air, glm-4.5-flash, glm-4.6, glm-4.7 from anthropic-messages to openai-completions API - Updated baseUrl from https://api.z.ai/api/anthropic to https://api.z.ai/api/coding/paas/v4 - Added compat setting to disable developer role for zai models - Filter empty text blocks in openai-completions to avoid zai API validation errors - Fixed zai provider tests to use OpenAI-style options (reasoningEffort)
This commit is contained in:
Anton Kuzmenko
parent
65fb9116fb
commit
0250b7ac03
4 changed files with 34 additions and 29 deletions
|
|
@ -258,7 +258,7 @@ async function loadModelsDevData(): Promise<Model<any>[]> {
|
|||
}
|
||||
}
|
||||
|
||||
// Process xAi models
|
||||
// Process zAi models
|
||||
if (data.zai?.models) {
|
||||
for (const [modelId, model] of Object.entries(data.zai.models)) {
|
||||
const m = model as ModelsDevModel;
|
||||
|
|
@ -268,9 +268,9 @@ async function loadModelsDevData(): Promise<Model<any>[]> {
|
|||
models.push({
|
||||
id: modelId,
|
||||
name: m.name || modelId,
|
||||
api: supportsImage ? "openai-completions" : "anthropic-messages",
|
||||
api: "openai-completions",
|
||||
provider: "zai",
|
||||
baseUrl: supportsImage ? "https://api.z.ai/api/coding/paas/v4" : "https://api.z.ai/api/anthropic",
|
||||
baseUrl: "https://api.z.ai/api/coding/paas/v4",
|
||||
reasoning: m.reasoning === true,
|
||||
input: supportsImage ? ["text", "image"] : ["text"],
|
||||
cost: {
|
||||
|
|
@ -279,11 +279,9 @@ async function loadModelsDevData(): Promise<Model<any>[]> {
|
|||
cacheRead: m.cost?.cache_read || 0,
|
||||
cacheWrite: m.cost?.cache_write || 0,
|
||||
},
|
||||
...(supportsImage ? {
|
||||
compat: {
|
||||
supportsDeveloperRole: false,
|
||||
},
|
||||
} : {}),
|
||||
compat: {
|
||||
supportsDeveloperRole: false,
|
||||
},
|
||||
contextWindow: m.limit?.context || 4096,
|
||||
maxTokens: m.limit?.output || 4096,
|
||||
});
|
||||
|
|
|
|||
|
|
@ -6978,9 +6978,10 @@ export const MODELS = {
|
|||
"glm-4.5": {
|
||||
id: "glm-4.5",
|
||||
name: "GLM-4.5",
|
||||
api: "anthropic-messages",
|
||||
api: "openai-completions",
|
||||
provider: "zai",
|
||||
baseUrl: "https://api.z.ai/api/anthropic",
|
||||
baseUrl: "https://api.z.ai/api/coding/paas/v4",
|
||||
compat: {"supportsDeveloperRole":false},
|
||||
reasoning: true,
|
||||
input: ["text"],
|
||||
cost: {
|
||||
|
|
@ -6991,13 +6992,14 @@ export const MODELS = {
|
|||
},
|
||||
contextWindow: 131072,
|
||||
maxTokens: 98304,
|
||||
} satisfies Model<"anthropic-messages">,
|
||||
} satisfies Model<"openai-completions">,
|
||||
"glm-4.5-air": {
|
||||
id: "glm-4.5-air",
|
||||
name: "GLM-4.5-Air",
|
||||
api: "anthropic-messages",
|
||||
api: "openai-completions",
|
||||
provider: "zai",
|
||||
baseUrl: "https://api.z.ai/api/anthropic",
|
||||
baseUrl: "https://api.z.ai/api/coding/paas/v4",
|
||||
compat: {"supportsDeveloperRole":false},
|
||||
reasoning: true,
|
||||
input: ["text"],
|
||||
cost: {
|
||||
|
|
@ -7008,13 +7010,14 @@ export const MODELS = {
|
|||
},
|
||||
contextWindow: 131072,
|
||||
maxTokens: 98304,
|
||||
} satisfies Model<"anthropic-messages">,
|
||||
} satisfies Model<"openai-completions">,
|
||||
"glm-4.5-flash": {
|
||||
id: "glm-4.5-flash",
|
||||
name: "GLM-4.5-Flash",
|
||||
api: "anthropic-messages",
|
||||
api: "openai-completions",
|
||||
provider: "zai",
|
||||
baseUrl: "https://api.z.ai/api/anthropic",
|
||||
baseUrl: "https://api.z.ai/api/coding/paas/v4",
|
||||
compat: {"supportsDeveloperRole":false},
|
||||
reasoning: true,
|
||||
input: ["text"],
|
||||
cost: {
|
||||
|
|
@ -7025,7 +7028,7 @@ export const MODELS = {
|
|||
},
|
||||
contextWindow: 131072,
|
||||
maxTokens: 98304,
|
||||
} satisfies Model<"anthropic-messages">,
|
||||
} satisfies Model<"openai-completions">,
|
||||
"glm-4.5v": {
|
||||
id: "glm-4.5v",
|
||||
name: "GLM-4.5V",
|
||||
|
|
@ -7047,9 +7050,10 @@ export const MODELS = {
|
|||
"glm-4.6": {
|
||||
id: "glm-4.6",
|
||||
name: "GLM-4.6",
|
||||
api: "anthropic-messages",
|
||||
api: "openai-completions",
|
||||
provider: "zai",
|
||||
baseUrl: "https://api.z.ai/api/anthropic",
|
||||
baseUrl: "https://api.z.ai/api/coding/paas/v4",
|
||||
compat: {"supportsDeveloperRole":false},
|
||||
reasoning: true,
|
||||
input: ["text"],
|
||||
cost: {
|
||||
|
|
@ -7060,7 +7064,7 @@ export const MODELS = {
|
|||
},
|
||||
contextWindow: 204800,
|
||||
maxTokens: 131072,
|
||||
} satisfies Model<"anthropic-messages">,
|
||||
} satisfies Model<"openai-completions">,
|
||||
"glm-4.6v": {
|
||||
id: "glm-4.6v",
|
||||
name: "GLM-4.6V",
|
||||
|
|
@ -7082,9 +7086,10 @@ export const MODELS = {
|
|||
"glm-4.7": {
|
||||
id: "glm-4.7",
|
||||
name: "GLM-4.7",
|
||||
api: "anthropic-messages",
|
||||
api: "openai-completions",
|
||||
provider: "zai",
|
||||
baseUrl: "https://api.z.ai/api/anthropic",
|
||||
baseUrl: "https://api.z.ai/api/coding/paas/v4",
|
||||
compat: {"supportsDeveloperRole":false},
|
||||
reasoning: true,
|
||||
input: ["text"],
|
||||
cost: {
|
||||
|
|
@ -7095,6 +7100,6 @@ export const MODELS = {
|
|||
},
|
||||
contextWindow: 204800,
|
||||
maxTokens: 131072,
|
||||
} satisfies Model<"anthropic-messages">,
|
||||
} satisfies Model<"openai-completions">,
|
||||
},
|
||||
} as const;
|
||||
|
|
|
|||
|
|
@ -460,13 +460,15 @@ function convertMessages(
|
|||
};
|
||||
|
||||
const textBlocks = msg.content.filter((b) => b.type === "text") as TextContent[];
|
||||
if (textBlocks.length > 0) {
|
||||
// Filter out empty text blocks to avoid API validation errors
|
||||
const nonEmptyTextBlocks = textBlocks.filter((b) => b.text && b.text.trim().length > 0);
|
||||
if (nonEmptyTextBlocks.length > 0) {
|
||||
// GitHub Copilot requires assistant content as a string, not an array.
|
||||
// Sending as array causes Claude models to re-answer all previous prompts.
|
||||
if (model.provider === "github-copilot") {
|
||||
assistantMsg.content = textBlocks.map((b) => sanitizeSurrogates(b.text)).join("");
|
||||
assistantMsg.content = nonEmptyTextBlocks.map((b) => sanitizeSurrogates(b.text)).join("");
|
||||
} else {
|
||||
assistantMsg.content = textBlocks.map((b) => {
|
||||
assistantMsg.content = nonEmptyTextBlocks.map((b) => {
|
||||
return { type: "text", text: sanitizeSurrogates(b.text) };
|
||||
});
|
||||
}
|
||||
|
|
|
|||
|
|
@ -556,7 +556,7 @@ describe("Generate E2E Tests", () => {
|
|||
});
|
||||
});
|
||||
|
||||
describe.skipIf(!process.env.ZAI_API_KEY)("zAI Provider (glm-4.5-air via Anthropic Messages)", () => {
|
||||
describe.skipIf(!process.env.ZAI_API_KEY)("zAI Provider (glm-4.5-air via OpenAI Completions)", () => {
|
||||
const llm = getModel("zai", "glm-4.5-air");
|
||||
|
||||
it("should complete basic text generation", { retry: 3 }, async () => {
|
||||
|
|
@ -572,11 +572,11 @@ describe("Generate E2E Tests", () => {
|
|||
});
|
||||
|
||||
it.skip("should handle thinking mode", { retry: 3 }, async () => {
|
||||
await handleThinking(llm, { thinkingEnabled: true, thinkingBudgetTokens: 2048 });
|
||||
await handleThinking(llm, { reasoningEffort: "medium" });
|
||||
});
|
||||
|
||||
it("should handle multi-turn with thinking and tools", { retry: 3 }, async () => {
|
||||
await multiTurn(llm, { thinkingEnabled: true, thinkingBudgetTokens: 2048 });
|
||||
await multiTurn(llm, { reasoningEffort: "medium" });
|
||||
});
|
||||
});
|
||||
|
||||
|
|
|
|||
Loading…
Add table
Add a link
Reference in a new issue