@@ -20815,11 +20815,30 @@ components:
2081520815 options:
2081620816 $ref: '#/components/schemas/ProviderOptions'
2081720817 type: object
20818+ response_format:
20819+ description: >-
20820+ Output format. "json" (default) returns { text, usage }. "verbose_json" additionally returns task, language,
20821+ duration, and segment-level timestamps; only supported by OpenAI-compatible providers.
20822+ enum:
20823+ - json
20824+ - verbose_json
20825+ example: json
20826+ type: string
2081820827 temperature:
2081920828 description: Sampling temperature for transcription
2082020829 example: 0
2082120830 format: double
2082220831 type: number
20832+ timestamp_granularities:
20833+ description: >-
20834+ Timestamp detail levels to include when response_format is "verbose_json". "segment" returns segment-level
20835+ timestamps; "word" additionally returns word-level timestamps in the words array. Ignored unless
20836+ response_format is "verbose_json".
20837+ example:
20838+ - segment
20839+ items:
20840+ $ref: '#/components/schemas/STTTimestampGranularity'
20841+ type: array
2082320842 required:
2082420843 - model
2082520844 - input_audio
@@ -20835,6 +20854,24 @@ components:
2083520854 seconds: 9.2
2083620855 total_tokens: 113
2083720856 properties:
20857+ duration:
20858+ description: Duration of the input audio in seconds, present when response_format is verbose_json
20859+ example: 9.2
20860+ format: double
20861+ type: number
20862+ language:
20863+ description: Detected or forced language, present when response_format is verbose_json
20864+ example: english
20865+ type: string
20866+ segments:
20867+ description: Timestamped transcript segments, present when response_format is verbose_json
20868+ items:
20869+ $ref: '#/components/schemas/STTSegment'
20870+ type: array
20871+ task:
20872+ description: The task performed, present when response_format is verbose_json
20873+ example: transcribe
20874+ type: string
2083820875 text:
2083920876 description: The transcribed text
2084020877 example: >-
@@ -20843,9 +20880,87 @@ components:
2084320880 type: string
2084420881 usage:
2084520882 $ref: '#/components/schemas/STTUsage'
20883+ words:
20884+ description: Timestamped words, present when the provider returns word-level timestamps
20885+ items:
20886+ $ref: '#/components/schemas/STTWord'
20887+ type: array
2084620888 required:
2084720889 - text
2084820890 type: object
20891+ STTSegment:
20892+ description: A timestamped transcript segment, returned when response_format is verbose_json
20893+ example:
20894+ avg_logprob: -0.28
20895+ compression_ratio: 1.13
20896+ end: 3.2
20897+ id: 0
20898+ no_speech_prob: 0.01
20899+ seek: 0
20900+ start: 0
20901+ temperature: 0
20902+ text: Hello there.
20903+ tokens:
20904+ - 50364
20905+ - 2425
20906+ - 456
20907+ properties:
20908+ avg_logprob:
20909+ description: Average log probability of the segment
20910+ format: double
20911+ type: number
20912+ compression_ratio:
20913+ description: Compression ratio of the segment
20914+ format: double
20915+ type: number
20916+ end:
20917+ description: Segment end time in seconds
20918+ example: 3.2
20919+ format: double
20920+ type: number
20921+ id:
20922+ description: Segment index within the transcript
20923+ example: 0
20924+ type: integer
20925+ no_speech_prob:
20926+ description: Probability the segment contains no speech
20927+ format: double
20928+ type: number
20929+ seek:
20930+ description: Seek offset of the segment
20931+ example: 0
20932+ type: integer
20933+ start:
20934+ description: Segment start time in seconds
20935+ example: 0
20936+ format: double
20937+ type: number
20938+ temperature:
20939+ description: Temperature used for the segment
20940+ format: double
20941+ type: number
20942+ text:
20943+ description: Transcribed text of the segment
20944+ example: Hello there.
20945+ type: string
20946+ tokens:
20947+ description: Token IDs of the segment
20948+ items:
20949+ type: integer
20950+ type: array
20951+ required:
20952+ - id
20953+ - start
20954+ - end
20955+ - text
20956+ type: object
20957+ STTTimestampGranularity:
20958+ description: A timestamp detail level for verbose_json transcription responses.
20959+ enum:
20960+ - word
20961+ - segment
20962+ example: word
20963+ type: string
2084920964 STTUsage:
2085020965 description: Aggregated usage statistics for the request
2085120966 example:
@@ -20878,6 +20993,32 @@ components:
2087820993 example: 113
2087920994 type: integer
2088020995 type: object
20996+ STTWord:
20997+ description: A timestamped word, returned when the provider includes word-level timestamps
20998+ example:
20999+ end: 0.4
21000+ start: 0
21001+ word: Hello
21002+ properties:
21003+ end:
21004+ description: Word end time in seconds
21005+ example: 0.4
21006+ format: double
21007+ type: number
21008+ start:
21009+ description: Word start time in seconds
21010+ example: 0
21011+ format: double
21012+ type: number
21013+ word:
21014+ description: The transcribed word
21015+ example: Hello
21016+ type: string
21017+ required:
21018+ - word
21019+ - start
21020+ - end
21021+ type: object
2088121022 SubagentNestedTool:
2088221023 additionalProperties:
2088321024 nullable: true
@@ -23932,13 +24073,26 @@ paths:
2393224073 description: The model to use for transcription.
2393324074 type: string
2393424075 response_format:
23935- description: The response format. Only "json" is supported.
24076+ description: >-
24077+ The response format. "json" (default) returns { text, usage }; "verbose_json" additionally returns
24078+ task, language, duration, and segment-level timestamps (OpenAI-compatible providers only).
2393624079 enum:
2393724080 - json
24081+ - verbose_json
2393824082 type: string
2393924083 temperature:
2394024084 description: The sampling temperature.
2394124085 type: number
24086+ timestamp_granularities[]:
24087+ description: >-
24088+ Timestamp detail levels to include when response_format is "verbose_json". "word" additionally
24089+ returns word-level timestamps in the words array.
24090+ items:
24091+ enum:
24092+ - word
24093+ - segment
24094+ type: string
24095+ type: array
2394224096 required:
2394324097 - file
2394424098 - model
@@ -25685,6 +25839,13 @@ paths:
2568525839 reserved permaslug `other` and is always returned last within its date, so callers
2568625840 can compute `top-50 traffic / total daily traffic` without a second request.
2568725841
25842+ Optional filters slice the dataset. `period` (`day`/`week`/`month`) sets the time
25843+ grain. `modality` and `context_bucket` narrow the exact dataset by output/input
25844+ modality (or tool-calling activity) and request context length. `category` and
25845+ `language_type` instead read a sampled, upsampled dataset whose `total_tokens` are
25846+ weekly-grain estimates — they cannot be combined with each other or with the exact
25847+ filters, and reject `period=day` with a 400.
25848+
2568825849 Authenticate with any valid OpenRouter API key (same key used for inference).
2568925850 Rate-limited to 30 requests/minute per key and 500 requests/day per account.
2569025851
@@ -25725,6 +25886,110 @@ paths:
2572525886 example: '2026-05-11'
2572625887 pattern: ^\d{4}-\d{2}-\d{2}$
2572725888 type: string
25889+ - description: >-
25890+ Time grain of each row. `day` (default) returns the per-UTC-day series; `week` buckets by ISO week start;
25891+ `month` buckets by month start. With `category` or `language_type` only `week` (default) and `month` are
25892+ available — `day` is rejected with a 400 because those datasets are aggregated weekly. For those sampled
25893+ datasets `period=month` buckets each week by its week-start month, so totals are approximate at month
25894+ boundaries.
25895+ in: query
25896+ name: period
25897+ required: false
25898+ schema:
25899+ description: >-
25900+ Time grain of each row. `day` (default) returns the per-UTC-day series; `week` buckets by ISO week start;
25901+ `month` buckets by month start. With `category` or `language_type` only `week` (default) and `month` are
25902+ available — `day` is rejected with a 400 because those datasets are aggregated weekly. For those sampled
25903+ datasets `period=month` buckets each week by its week-start month, so totals are approximate at month
25904+ boundaries.
25905+ enum:
25906+ - day
25907+ - week
25908+ - month
25909+ example: day
25910+ type: string
25911+ - description: >-
25912+ Restrict to models for a modality surface: `text` / `image_output` match output modality, `image` / `audio`
25913+ match input modality, and `tool_calling` keeps only rows that recorded at least one tool call. Exact dataset
25914+ — cannot be combined with `category` or `language_type`.
25915+ in: query
25916+ name: modality
25917+ required: false
25918+ schema:
25919+ description: >-
25920+ Restrict to models for a modality surface: `text` / `image_output` match output modality, `image` /
25921+ `audio` match input modality, and `tool_calling` keeps only rows that recorded at least one tool call.
25922+ Exact dataset — cannot be combined with `category` or `language_type`.
25923+ enum:
25924+ - text
25925+ - image
25926+ - image_output
25927+ - audio
25928+ - tool_calling
25929+ example: text
25930+ type: string
25931+ - description: >-
25932+ Restrict to requests whose context length falls in this bucket (`1K`, `10K`, `100K`, `1M`, or `10M`). Exact
25933+ dataset — cannot be combined with `category` or `language_type`.
25934+ in: query
25935+ name: context_bucket
25936+ required: false
25937+ schema:
25938+ description: >-
25939+ Restrict to requests whose context length falls in this bucket (`1K`, `10K`, `100K`, `1M`, or `10M`).
25940+ Exact dataset — cannot be combined with `category` or `language_type`.
25941+ enum:
25942+ - 1K
25943+ - 10K
25944+ - 100K
25945+ - 1M
25946+ - 10M
25947+ example: 100K
25948+ type: string
25949+ - description: >-
25950+ Restrict to a use-case category (e.g. `programming`, `roleplay`). Sourced from a sampled, upsampled dataset,
25951+ so `total_tokens` is an estimate and is aggregated weekly (the trailing weekly bucket may include traffic
25952+ past `end_date`). Cannot be combined with `modality`, `context_bucket`, or `language_type`.
25953+ in: query
25954+ name: category
25955+ required: false
25956+ schema:
25957+ description: >-
25958+ Restrict to a use-case category (e.g. `programming`, `roleplay`). Sourced from a sampled, upsampled
25959+ dataset, so `total_tokens` is an estimate and is aggregated weekly (the trailing weekly bucket may include
25960+ traffic past `end_date`). Cannot be combined with `modality`, `context_bucket`, or `language_type`.
25961+ enum:
25962+ - programming
25963+ - roleplay
25964+ - marketing
25965+ - marketing/seo
25966+ - technology
25967+ - science
25968+ - translation
25969+ - legal
25970+ - finance
25971+ - health
25972+ - trivia
25973+ - academia
25974+ example: programming
25975+ type: string
25976+ - description: >-
25977+ Restrict to natural-language or programming-language tagged activity. Sourced from a sampled, upsampled
25978+ dataset, so `total_tokens` is an estimate and is aggregated weekly (the trailing weekly bucket may include
25979+ traffic past `end_date`). Cannot be combined with `modality`, `context_bucket`, or `category`.
25980+ in: query
25981+ name: language_type
25982+ required: false
25983+ schema:
25984+ description: >-
25985+ Restrict to natural-language or programming-language tagged activity. Sourced from a sampled, upsampled
25986+ dataset, so `total_tokens` is an estimate and is aggregated weekly (the trailing weekly bucket may include
25987+ traffic past `end_date`). Cannot be combined with `modality`, `context_bucket`, or `category`.
25988+ enum:
25989+ - natural
25990+ - programming
25991+ example: natural
25992+ type: string
2572825993 responses:
2572925994 '200':
2573025995 content:
0 commit comments