Skip to content

Text Model Matrix ​

This page organizes the text series API by protocol and task type. Model names are based on MoonNexAI console and API Reference, using the available models in request body model field.

Protocol and entrance ​

sceneRecommended entranceCommon ModelsReturns byFit to scene
Dialogue CompletionPOST /v1/chat/completionsgpt-5.6-terrasynchronous or streamingChat client, assistant, workstream node.
ResponsesPOST /v1/responsesgpt-5.6-terrasynchronous or streamingMulti-modular input, tool call style and context management.
Respondes Context CompressionPOST /v1/responses/compactModels that support context compressionsynchronousLong dialogue is constricted, key context is preserved and follow-up request is reduced.
MessagesPOST /v1/messagesmoonnexai-claudesynchronous or streamingClaude compatible client.
Text embeddingsPOST /v1/embeddingsmoonnexai-embeddingsynchronousSearch, sort, group and semantic match.
Text transcribingPOST /v1/audio/speechtts-1, tts-1-hdsynchronousVoice broadcasts, voice-speeching and content reading.
Audio transcriptionPOST /v1/audio/transcriptionswhisper-1synchronousConference, audio, video subtitle extraction.
Audio translationPOST /v1/audio/translationswhisper-1synchronousTranslation of audio content into target languages.

Common Models ​

ModelsPurposeCommon API
gpt-5.6-terraDialogue, Responses and Common Text Generation./v1/chat/completions, /v1/responses
moonnexai-claudeMessages format text generation./v1/messages
moonnexai-embeddingText embeddings./v1/embeddings
tts-1Standard text transject./v1/audio/speech
tts-1-hdHigher quality text transcribing./v1/audio/speech
whisper-1transcription and translation./v1/audio/transcriptions, /v1/audio/translations

Select Recommendations ​

RequirementsRecommendations
OpenAI SDK already usedSet https://moonnexai.com/v1 with /v1/chat/completions or /v1/responses, Base URL.
A client with Claude compatible already is usedUse /v1/messages and confirm that the client supports custom Base URLs.
Semantic search requiredUse /v1/embeddings, save embeddings and original IDs on the operating system.
Voice input output required/v1/audio/speech, transcription and the translation use the corresponding audio API.

Relevant Pages ​