Skip to content

Enhanced ChatGPT Clone: Features Agents, MCP, Skills, DeepSeek, Anthropic, AWS, OpenAI, Responses API, Azure, Groq, o1, GPT-5, Mistral, OpenRouter, Vertex AI, Gemini, Artifacts, AI model switching, message search, Code Interpreter, langchain, DALL-E-3, OpenAPI Actions, Functions, Secure Multi-User Auth, Presets, open-source for self-hosting. Active

🧠 fix(Cohere): map to expected SDK params (#2329)

Danny Avilafb80af05be69
diff --git a/api/app/clients/llm/createCoherePayload.js b/api/app/clients/llm/createCoherePayload.js
index d399f7055a..58803d76f3 100644
--- a/api/app/clients/llm/createCoherePayload.js
+++ b/api/app/clients/llm/createCoherePayload.js
@@ -69,13 +69,13 @@ function createCoherePayload({ modelOptions }) {
return {
message: latestUserMessageContent,
model: model,
- chat_history: chatHistory,
+ chatHistory,
stream: stream ?? false,
temperature: temperature,
- frequency_penalty: frequency_penalty,
- presence_penalty: presence_penalty,
- max_tokens: max_tokens,
- stop_sequences: stop,
+ frequencyPenalty: frequency_penalty,
+ presencePenalty: presence_penalty,
+ maxTokens: max_tokens,
+ stopSequences: stop,
preamble,
p: top_p,
...rest,
diff --git a/docs/install/configuration/ai_endpoints.md b/docs/install/configuration/ai_endpoints.md
index d6142c7f91..787f43898b 100644
--- a/docs/install/configuration/ai_endpoints.md
+++ b/docs/install/configuration/ai_endpoints.md
@@ -23,13 +23,13 @@ Some of the endpoints are marked as **Known,** which means they might have speci
- Experimental: does not follow OpenAI-spec, uses a new method for endpoint compatibility, shares some similarities and parameters.
- For a full list of Cohere-specific parameters, see the [Cohere API documentation](https://docs.cohere.com/reference/chat).
- Note: The following parameters are recognized between OpenAI and Cohere. Most are removed in the example config below to prefer Cohere's default settings:
- - `stop`: mapped to `stop_sequences`
+ - `stop`: mapped to `stopSequences`
- `top_p`: mapped to `p`, different min/max values
- - `frequency_penalty`: different min/max values
- - `presence_penalty`: different min/max values
+ - `frequency_penalty`: mapped to `frequencyPenalty`, different min/max values
+ - `presence_penalty`: mapped to `presencePenalty`, different min/max values
- `model`: shared, included by default.
- `stream`: shared, included by default.
- - `max_tokens`: shared, not included by default.
+ - `max_tokens`: shared, mapped to `maxTokens`, not included by default.
```yaml