Skip to content
Documentation
GuidesAPI ReferenceIntegrations
ArenaPlatform
API Reference/Text to Speech

Generate Speech

POST/v1/audio/speech

Overview

Generate speech with an exact provider model or tts/auto.

Routing and provider details

Responses stream binary audio unless a supported provider transport explicitly returns JSON. Native voice IDs are accepted for exact models; tts/auto chooses a portable voice. Model-specific formats, languages and limits are exposed in the catalog. Published provider tariffs minus 15% apply before account credits. Reservations precede provider calls and native completion settles usage.

MiniMax speech-2.8, speech-2.6 and speech-02 HD/Turbo support background jobs using the -async model suffix and ?async=true, up to 50,000 characters. The 202 response contains the request ID. Poll GET /v1/audio/speech/{id} with the same API key to download the completed result; it expires after 24 hours. Async requests require a healthy worker and workspace audio storage.

Azure neural and neural-hd-flash accept their corresponding native voice IDs. Alibaba CosyVoice 3.5 requires a provisioned custom voice and a separately configured Beijing credential; Singapore credentials are not interchangeable.

Authorization

Authorizationstring · headerRequired

AudioLab API key sent as Bearer <token>. Authentication guide

Parameters

asyncquery · string

Queue a supported MiniMax -async model instead of waiting.

Request Body

application/json
modelstringRequired

Exact provider/model slug or a managed TTS alias. Preview, mutable, compatibility-only, and estimated-price models require exact selection unless tts/auto feature routing explicitly ranks them.

featurestring

Case-sensitive TTS routing capability. Valid only with model="tts/auto"; eligible models are tried in feature rank order.

Values: "Acting""Expressiveness""Voice identity""Language stability""Reliability""Extended long-form""Long-form""Acoustic quality"

inputstringRequired
voicestring | objectRequired
Show nested fields
Option 1: string

Provider-native voice ID/name, or auto where the selected model has a default. resemble/chatterbox-turbo requires a custom DeepInfra voice ID and rejects auto.

Option 2: object
idstringRequired
response_formatstring

Supported values are model-specific; inspect the selected model's endpoint limits. Gemini 2.5 TTS accepts only wav or pcm.

Values: "mp3""ogg""opus""aac""flac""wav""pcm""ulaw""alaw""json"

Default: "mp3"

speednumber

Default: 1

Minimum 0.25 · Maximum 4

instructionsstring
languagestring
providerobject
Show nested fields
orderstring[]
Show nested fields

string

onlystring[]
Show nested fields

string

ignorestring[]
Show nested fields

string

modelsstring[]

Model ID patterns allowed for managed routes. Supports * wildcards.

Show nested fields

string

sortstring

Values: "price""latency""quality""balanced"

cost_qualityinteger

Balanced-sort weighting from quality (0) to cost (10).

Minimum 0 · Maximum 10

allow_fallbacksboolean
max_priceobject
Show nested fields
per_1k_charsnumber

Minimum 0

data_regionstring | string[]

Route only to endpoints documented as processing in this region: us, eu (which includes EU member states), or a country code such as nl, sg, in or cn. A list matches any of its regions. Endpoints whose region is global or unknown never match; global on its own means no requirement. The request's region narrows the workspace's Data Region and a named router's region policy, and never widens them. Fallbacks keep to it. When no endpoint qualifies, the request is refused with 400 data_region_unavailable, param provider.data_region.

Show nested fields
Option 1: string

string

Option 2: string[]

See the OpenAPI schema for this nested object.

regionstring[]

The older name of data_region, with the same meaning.

Show nested fields

string

optionsobject
Show nested fields
alibabaobject
Show nested fields
transportstring

HTTP only. File returns Alibaba's hosted WAV; SSE streams base64 24 kHz PCM, which AudioLab can return directly or transcode.

Values: "file""sse"

language_typestring

Values: "Auto""Chinese""English""German""Italian""Portuguese""Spanish""Japanese""Korean""French""Russian"

optimize_instructionsboolean

Qwen3 TTS Instruct models only. Requires common instructions and asks Alibaba to rewrite them for natural delivery.

modestring

WebSocket only. Server commit balances latency and quality; commit requires explicit input_text_buffer.commit events.

Values: "server_commit""commit"

Default: "server_commit"

sample_rateinteger

WebSocket only.

Values: 8000160002400048000

Default: 24000

speech_ratenumber

WebSocket only. Overrides common speed.

Minimum 0.5 · Maximum 2

volumeinteger

WebSocket only.

Minimum 0 · Maximum 100

pitch_ratenumber

WebSocket only.

Minimum 0.5 · Maximum 2

bit_rateinteger

WebSocket Opus output only, in kbps.

Minimum 6 · Maximum 510

openaiobject
Show nested fields
stream_formatstring

Native OpenAI stream envelope. SSE is available only on GPT-4o mini TTS models and is passed through as text/event-stream.

Values: "audio""sse"

Default: "audio"

amazonobject
Show nested fields
language_codestring

Values: "arb""cmn-CN""cy-GB""da-DK""de-DE""en-AU""en-GB""en-GB-WLS""en-IN""en-US""es-ES""es-MX""es-US""fr-CA""fr-FR""is-IS""it-IT""ja-JP""hi-IN""ko-KR""nb-NO""nl-NL""pl-PL""pt-BR""pt-PT""ro-RO""ru-RU""sv-SE""tr-TR""en-NZ""en-ZA""ca-ES""de-AT""yue-CN""ar-AE""fi-FI""en-IE""nl-BE""fr-BE""cs-CZ""de-CH""en-SG"

lexicon_namesstring[]
Show nested fields

See the OpenAPI schema for this nested object.

sample_rateinteger

Valid values depend on response_format. Opus is fixed at 48000; μ-law and A-law at 8000; PCM permits 8000 or 16000.

Values: 80001600022050240004410048000

speech_mark_typesstring[]

Requires response_format=json and is unavailable on the generative engine.

Show nested fields

See the OpenAPI schema for this nested object.

text_typestring

Values: "text""ssml"

Default: "text"

elevenlabsobject
Show nested fields
transportstring

Defaults to stream, except WAV defaults to create because ElevenLabs streaming endpoints do not expose WAV.

Values: "create""create_with_timing""stream""stream_with_timing"

enable_loggingboolean

False requests ElevenLabs zero-retention mode; Enterprise only.

optimize_streaming_latencyinteger | null

Minimum 0 · Maximum 4

output_formatstring

Values: "alaw_8000""mp3_22050_32""mp3_24000_48""mp3_44100_32""mp3_44100_64""mp3_44100_96""mp3_44100_128""mp3_44100_192""opus_48000_32""opus_48000_64""opus_48000_96""opus_48000_128""opus_48000_192""pcm_8000""pcm_16000""pcm_22050""pcm_24000""pcm_32000""pcm_44100""pcm_48000""ulaw_8000""wav_8000""wav_16000""wav_22050""wav_24000""wav_32000""wav_44100""wav_48000"

language_codestring | null
voice_settingsobject | null
Show nested fields

See the OpenAPI schema for this nested object.

pronunciation_dictionary_locatorsarray | null
Show nested fields

See the OpenAPI schema for this nested object.

seedinteger | null

Minimum 0 · Maximum 4294967295

previous_textstring | null
next_textstring | null
previous_request_idsarray | null
Show nested fields

See the OpenAPI schema for this nested object.

next_request_idsarray | null
Show nested fields

See the OpenAPI schema for this nested object.

use_pvc_as_ivcboolean
apply_text_normalizationstring

Values: "auto""on""off"

apply_language_text_normalizationboolean

Currently supported by ElevenLabs for Japanese.

stabilitynumber

AudioLab compatibility alias; prefer voice_settings.stability.

Minimum 0 · Maximum 1

similarity_boostnumber

AudioLab compatibility alias; prefer voice_settings.similarity_boost.

Minimum 0 · Maximum 1

stylenumber

AudioLab compatibility alias; prefer voice_settings.style.

Minimum 0 · Maximum 1

use_speaker_boostboolean

AudioLab compatibility alias; prefer voice_settings.use_speaker_boost.

cartesiaobject
Show nested fields
transportstring

Values: "bytes""sse"

Default: "bytes"

output_formatobject | object | object
Show nested fields

See the OpenAPI schema for this nested object.

languagestring | null
Show nested fields

See the OpenAPI schema for this nested object.

pronunciation_dict_idstring | null

Sonic 3 and newer, including Sonic 3.5.

generation_configobject
Show nested fields

See the OpenAPI schema for this nested object.

context_idstring | null

SSE only.

add_timestampsboolean | null

SSE only.

add_phoneme_timestampsboolean | null

SSE only.

use_normalized_timestampsboolean | null

SSE only.

deepinfraobject
Show nested fields
transportstring

OpenAI-compatible raw audio, native metadata JSON, Kokoro native streaming, or DeepInfra's buffered/streaming ElevenLabs-compatible route.

Values: "openai""native_json""native_stream""elevenlabs""elevenlabs_stream"

Default: "openai"

service_tierstring

OpenAI and native transports only.

Values: "default""priority""flex"

versionstring

Native transports only; selects a specific DeepInfra model version.

language_codestring

ISO 639-1 code for ElevenLabs-compatible transports; must match common language when both are supplied.

webhookstring

native_json only; asks DeepInfra to notify this URL when inference completes.

kokoroobject
Show nested fields

See the OpenAPI schema for this nested object.

chatterboxobject
Show nested fields

See the OpenAPI schema for this nested object.

deepgramobject
Show nested fields
encodingstring

Native output encoding. It must produce the common response_format together with container.

Values: "linear16""mulaw""alaw""mp3""opus""flac""aac"

containerstring

wav/none for linear16, mulaw, and alaw; ogg for Opus; omitted for MP3, FLAC, and AAC.

Values: "none""wav""ogg"

sample_rateinteger

Configurable only for linear16, mu-law, A-law, and FLAC. MP3, Opus, and AAC use fixed provider rates and AudioLab omits this query parameter.

Values: 80001600022050240003200048000

bit_rateinteger

MP3 accepts 32000/48000, Opus accepts 4000-650000, and AAC accepts 4000-192000.

Minimum 4000 · Maximum 650000

mip_opt_outboolean

Excludes this request from Deepgram's Model Improvement Program without a PAYG/Growth rate change as of March 5, 2026.

Default: false

tagstring | string[]

One or more immutable Deepgram usage-reporting tags.

Show nested fields

See the OpenAPI schema for this nested object.

callbackstring

Asynchronously sends generated audio to an HTTP(S) URL on port 80, 443, 8080, or 8443; the AudioLab response is Deepgram's immediate JSON request ID.

callback_methodstring

Requires callback.

Values: "POST""PUT""post""put"

Default: "POST"

geminiobject
Show nested fields
transportstring

Gemini 3.1 supports all four transports. Gemini 2.5 supports unary interactions and generate_content only.

Values: "interactions""interactions_stream""generate_content""stream_generate_content"

Default: "interactions_stream"

speech_configobject[]
Show nested fields

See the OpenAPI schema for this nested object.

response_formatobject
Show nested fields

See the OpenAPI schema for this nested object.

storeboolean

Interactions only. Explicitly opt in to Google's interaction storage.

Default: false

max_output_tokensinteger

Minimum 1 · Maximum 16384

seedinteger

Minimum 0 · Maximum 4294967295

stop_sequencesstring[]
Show nested fields

See the OpenAPI schema for this nested object.

temperaturenumber

Minimum 0 · Maximum 2

top_pnumber

Minimum 0 · Maximum 1

humeobject
Show nested fields
transportstring

Values: "file""json""stream_file""stream_json"

Default: "stream_file"

versionstring

Omit to let Hume automatically select the Octave version.

Values: "1""2"

voice_providerstring

AudioLab mapping hint for the common voice field. Names default to HUME_AI. HTTP voice IDs first use CUSTOM_VOICE and retry a zero-byte pre-audio response as HUME_AI. Set this field to pin the mapping; WebSocket voice IDs should always set it explicitly.

Values: "HUME_AI""CUSTOM_VOICE"

contextobject | object
Show nested fields

See the OpenAPI schema for this nested object.

include_timestamp_typesstring[]

Octave 2 only; requires json or stream_json so timestamp metadata is preserved.

Show nested fields

See the OpenAPI schema for this nested object.

num_generationsinteger

Minimum 1 · Maximum 5

split_utterancesboolean

Default: true

strip_headersboolean

Default: false

temperaturenumber | null

Minimum 0.1 · Maximum 1

instant_modeboolean

Streaming transports only. Requires a voice and exactly one generation.

trailing_silencenumber

Applied to the generated utterance.

Minimum 0

inworldobject
Show nested fields
transportstring

Decoded buffered audio, decoded HTTP streaming audio, native JSON, or native newline-delimited streaming JSON.

Values: "buffered""stream""json""stream_json"

Default: "buffered"

audioConfigobject
Show nested fields

See the OpenAPI schema for this nested object.

temperaturenumber

TTS 1.5 Max and Mini only. Zero asks Inworld to use its model default; TTS-2 ignores this field, so AudioLab rejects it there.

Minimum 0 · Maximum 2

deliveryModestring

TTS-2 only. Controls output stability and variation; TTS 1.5 ignores this field, so AudioLab rejects it there.

Values: "DELIVERY_MODE_UNSPECIFIED""STABLE""BALANCED""CREATIVE"

timestampTypestring

WORD and CHARACTER require json or stream_json so alignment metadata is preserved.

Values: "TIMESTAMP_TYPE_UNSPECIFIED""WORD""CHARACTER"

applyTextNormalizationstring

Values: "APPLY_TEXT_NORMALIZATION_UNSPECIFIED""ON""OFF"

timestampTransportStrategystring

Requires stream_json and a WORD or CHARACTER timestampType.

Values: "TIMESTAMP_TRANSPORT_STRATEGY_UNSPECIFIED""SYNC""ASYNC"

minimaxobject
Show nested fields
transportstring

Decoded buffered audio, decoded SSE audio, native JSON, or native SSE.

Values: "buffered""stream""json""sse"

Default: "buffered"

stream_optionsobject

stream or sse only. Decoded stream defaults this to true to prevent duplicate final audio.

Show nested fields

See the OpenAPI schema for this nested object.

voice_settingobject
Show nested fields

See the OpenAPI schema for this nested object.

audio_settingobject
Show nested fields

See the OpenAPI schema for this nested object.

pronunciation_dictobject
Show nested fields

See the OpenAPI schema for this nested object.

timbre_weightsobject[]

Requires common voice=auto.

Show nested fields

See the OpenAPI schema for this nested object.

language_booststring

Values: "Chinese""Chinese,Yue""English""Arabic""Russian""Spanish""French""Portuguese""German""Turkish""Dutch""Ukrainian""Vietnamese""Indonesian""Japanese""Italian""Korean""Thai""Polish""Romanian""Greek""Czech""Finnish""Hindi""Bulgarian""Danish""Hebrew""Malay""Persian""Slovak""Swedish""Croatian""Filipino""Hungarian""Norwegian""Slovenian""Catalan""Nynorsk""Tamil""Afrikaans""auto"

voice_modifyobject
Show nested fields

See the OpenAPI schema for this nested object.

subtitle_enableboolean

Requires native json or sse so subtitle metadata is preserved.

Default: false

subtitle_typestring

word_streaming requires stream or sse.

Values: "sentence""word""word_streaming"

Default: "sentence"

output_formatstring

url is non-streaming, requires json, and is valid upstream for 24 hours.

Values: "hex""url"

Default: "hex"

murfobject
Show nested fields
transportstring

stream is available on both models. generate and json use Gen2 Synthesize Speech; json requires response_format=json and preserves Murf metadata.

Values: "stream""generate""json"

Default: "stream"

regionstring

Streaming HTTP and WebSocket host. Omit to use MURF_STREAM_BASE_URL.

Values: "global""us-east""us-west""in""ca""kr""me""jp""au""eu-central""uk""sa-east"

formatstring

Audio encoding embedded in a native JSON response. Other transports derive this from response_format.

Values: "MP3""FLAC""WAV""OGG""PCM""ULAW""ALAW"

channel_typestring

Values: "MONO""STEREO"

Default: "MONO"

sample_rateinteger

Synthesize Speech does not accept 16000 Hz. Falcon 2 defaults to 24000; Gen2 defaults to 44100.

Values: 800016000240004410048000

localestring

Voice-supported locale from Murf's live voice catalog. Murf uses en-UK for British English.

pitchinteger

Minimum -50 · Maximum 50

rateinteger

Native percentage rate. Overrides the common speed mapping.

Minimum -50 · Maximum 50

stylestring
variationinteger

Gen2 only.

Minimum 0 · Maximum 5

audio_durationnumber

Gen2 generate/json only; requests an approximate output duration in seconds.

Minimum 0

encode_as_base64boolean

Gen2 generate/json only. Inline Base64 enables Murf's zero-retention mode; AudioLab defaults this to true for generate audio.

word_durations_as_original_textboolean

Gen2 generate/json only; English inputs only.

mistralobject
Show nested fields
ref_audiostring

Standard base64-encoded one-off reference audio, without a data-URL prefix. Use this while common voice is auto; do not combine it with a saved voice ID.

xaiobject
Show nested fields
output_formatobject
Show nested fields

See the OpenAPI schema for this nested object.

optimize_streaming_latencyinteger

0 preserves quality; 1 reduces first-chunk size for lower time-to-first-audio.

Minimum 0 · Maximum 1

text_normalizationboolean

Default: false

with_timestampsboolean

REST returns XaiTiming JSON; WebSocket audio.delta events include timing fields.

Default: false

metadataobject

Responses

200Generated audio stream. Content-Type reflects response_format.
application/octet-stream

string

application/json
Option 1: object
audio_base64stringRequired
alignmentobject | null
Show nested fields
charactersstring[]
Show nested fields

See the OpenAPI schema for this nested object.

character_start_times_secondsnumber[]
Show nested fields

See the OpenAPI schema for this nested object.

character_end_times_secondsnumber[]
Show nested fields

See the OpenAPI schema for this nested object.

normalized_alignmentobject | null
Show nested fields
charactersstring[]
Show nested fields

See the OpenAPI schema for this nested object.

character_start_times_secondsnumber[]
Show nested fields

See the OpenAPI schema for this nested object.

character_end_times_secondsnumber[]
Show nested fields

See the OpenAPI schema for this nested object.

Option 2: object
audiostringRequired

Provider-encoded audio returned by native_json.

input_character_lengthinteger
output_formatstring
wordsobject[]
Show nested fields
idintegerRequired
startnumberRequired
endnumberRequired
textstringRequired
request_idstring
inference_statusobject
Show nested fields
statusstring

Values: "unknown""queued""running""succeeded""failed"

runtime_msinteger
costnumber
tokens_generatedinteger | null
tokens_inputinteger | null
output_lengthinteger | null
Option 3: object
request_idstringRequired

Deepgram request ID for asynchronous callback generation.

Option 4: object
generationsobject[]Required
Show nested fields
audiostringRequired
durationnumberRequired
encodingobjectRequired
Show nested fields

See the OpenAPI schema for this nested object.

file_sizeintegerRequired
generation_idstringRequired
snippetsobject[][]Required
Show nested fields

See the OpenAPI schema for this nested object.

request_idstring | nullRequired
Option 5: object
audioContentstring
usageobject
Show nested fields
processedCharactersCountinteger

Minimum 0

modelIdstring
timestampInfoobject
Show nested fields
wordAlignmentobject
Show nested fields
wordsstring[]
Show nested fields

See the OpenAPI schema for this nested object.

wordStartTimeSecondsnumber[]
Show nested fields

See the OpenAPI schema for this nested object.

wordEndTimeSecondsnumber[]
Show nested fields

See the OpenAPI schema for this nested object.

phoneticDetailsobject[]
Show nested fields

See the OpenAPI schema for this nested object.

characterAlignmentobject
Show nested fields
charactersstring[]
Show nested fields

See the OpenAPI schema for this nested object.

characterStartTimeSecondsnumber[]
Show nested fields

See the OpenAPI schema for this nested object.

characterEndTimeSecondsnumber[]
Show nested fields

See the OpenAPI schema for this nested object.

Option 6: object
dataobject | null
Show nested fields
audiostring

Hex audio or a hosted URL according to output_format.

subtitle_filestring
statusinteger

Values: 12

trace_idstring
extra_infoobject
base_respobject
Show nested fields
status_codeinteger
status_msgstring
Option 7: object
audioFilestring
encodedAudiostring

Base64 audio when encode_as_base64 is enabled.

audioLengthInSecondsnumber
remainingCharacterCountinteger
wordDurationsobject[]
Show nested fields
wordstring
startMsinteger
endMsinteger
warningstring
Option 8: object
audiostringRequired
content_typestringRequired
durationnumberRequired
audio_timestampsobjectRequired
Show nested fields
graph_charsstring[]Required
Show nested fields

See the OpenAPI schema for this nested object.

graph_timesnumber[][]Required
Show nested fields

See the OpenAPI schema for this nested object.

Option 9: string

Chunked ElevenLabs JSON timing objects from stream_with_timing.

text/event-stream

Native OpenAI, Cartesia, or MiniMax SSE audio, metadata, timestamp, completion, and error events.

application/x-json-stream

Amazon Polly sentence, SSML, viseme, or word speech marks, one JSON object per line.

application/x-ndjson

Native Hume or Inworld streamed-JSON audio, timestamp, usage, and status objects, one JSON object per line.

202Background speech job queued. Poll the returned id.
application/json
idstringRequired
statusstringRequired

Values: "queued"

400Request validation failed.
application/json
errorobjectRequired
Show nested fields
codestringRequired
messagestringRequired
paramstring | null
request_idstring | null
metadataobject

Typed detail for some refusals. budget_exceeded (402): limit (key_budget or workspace_budget) and budget with scope, period, limit, usage and resets_at. key_limit_exceeded (402): limit key_spend_limit. access_denied (403): policy (workspace or key), rule (models, providers or products), product, provider and model.

401Authentication is missing or invalid.
application/json
errorobjectRequired
Show nested fields
codestringRequired
messagestringRequired
paramstring | null
request_idstring | null
metadataobject

Typed detail for some refusals. budget_exceeded (402): limit (key_budget or workspace_budget) and budget with scope, period, limit, usage and resets_at. key_limit_exceeded (402): limit key_spend_limit. access_denied (403): policy (workspace or key), rule (models, providers or products), product, provider and model.

402Credits or billing setup is required.
application/json
errorobjectRequired
Show nested fields
codestringRequired
messagestringRequired
paramstring | null
request_idstring | null
metadataobject

Typed detail for some refusals. budget_exceeded (402): limit (key_budget or workspace_budget) and budget with scope, period, limit, usage and resets_at. key_limit_exceeded (402): limit key_spend_limit. access_denied (403): policy (workspace or key), rule (models, providers or products), product, provider and model.

403Authenticated principal lacks permission.
application/json
errorobjectRequired
Show nested fields
codestringRequired
messagestringRequired
paramstring | null
request_idstring | null
metadataobject

Typed detail for some refusals. budget_exceeded (402): limit (key_budget or workspace_budget) and budget with scope, period, limit, usage and resets_at. key_limit_exceeded (402): limit key_spend_limit. access_denied (403): policy (workspace or key), rule (models, providers or products), product, provider and model.

415Audio format is unsupported.
application/json
errorobjectRequired
Show nested fields
codestringRequired
messagestringRequired
paramstring | null
request_idstring | null
metadataobject

Typed detail for some refusals. budget_exceeded (402): limit (key_budget or workspace_budget) and budget with scope, period, limit, usage and resets_at. key_limit_exceeded (402): limit key_spend_limit. access_denied (403): policy (workspace or key), rule (models, providers or products), product, provider and model.

422Audio could not be processed.
application/json
errorobjectRequired
Show nested fields
codestringRequired
messagestringRequired
paramstring | null
request_idstring | null
metadataobject

Typed detail for some refusals. budget_exceeded (402): limit (key_budget or workspace_budget) and budget with scope, period, limit, usage and resets_at. key_limit_exceeded (402): limit key_spend_limit. access_denied (403): policy (workspace or key), rule (models, providers or products), product, provider and model.

429Rate limit exceeded.
application/json
errorobjectRequired
Show nested fields
codestringRequired
messagestringRequired
paramstring | null
request_idstring | null
metadataobject

Typed detail for some refusals. budget_exceeded (402): limit (key_budget or workspace_budget) and budget with scope, period, limit, usage and resets_at. key_limit_exceeded (402): limit key_spend_limit. access_denied (403): policy (workspace or key), rule (models, providers or products), product, provider and model.

500Internal server error.
application/json
errorobjectRequired
Show nested fields
codestringRequired
messagestringRequired
paramstring | null
request_idstring | null
metadataobject

Typed detail for some refusals. budget_exceeded (402): limit (key_budget or workspace_budget) and budget with scope, period, limit, usage and resets_at. key_limit_exceeded (402): limit key_spend_limit. access_denied (403): policy (workspace or key), rule (models, providers or products), product, provider and model.

502Upstream provider failed after allowed attempts.
application/json
errorobjectRequired
Show nested fields
codestringRequired
messagestringRequired
paramstring | null
request_idstring | null
metadataobject

Typed detail for some refusals. budget_exceeded (402): limit (key_budget or workspace_budget) and budget with scope, period, limit, usage and resets_at. key_limit_exceeded (402): limit key_spend_limit. access_denied (403): policy (workspace or key), rule (models, providers or products), product, provider and model.

503No eligible provider endpoint is currently available, or async processing has no healthy worker polling the durable queue.
application/json
errorobjectRequired
Show nested fields
codestringRequired
messagestringRequired
paramstring | null
request_idstring | null
metadataobject

Typed detail for some refusals. budget_exceeded (402): limit (key_budget or workspace_budget) and budget with scope, period, limit, usage and resets_at. key_limit_exceeded (402): limit key_spend_limit. access_denied (403): policy (workspace or key), rule (models, providers or products), product, provider and model.