# List models

Every model the gateway serves right now, with the ID to use in requests, accepted inputs
(text, images, video, ...), context length, supported parameters and the prices in EUR per
token of its currently cheapest route (input plus output token; tariff variants, which
serve only requests naming them, are left out unless the model has no other route). The selected route's
complete base tariff and conditional overrides are published together; unit fees are
per request/image/search/page/Unicode character, audio_second per started second. Same format as OpenRouter's
catalog, plus `pricing.currency`. Returns text-output models by default; use
`output_modalities=all` for the full catalog. No API key needed.

Providers with prices in another currency are converted with the gateway's current
exchange rate; `pricing.original` then shows their own prices and the rate. Routes
without a usable exchange rate cannot serve requests and are left out.

Aliases of the gateway, such as `eu/fast` (current recommended version) and
`eu/fast@2026-10` (pinned), are listed as entries of their own with the data of the model
they resolve to, which `canonical_slug` names. They are not part of OpenRouter's catalog.

`GET https://api.tokenlane.de/v1/models`

## Authorizations

[]

## curl

```curl
curl --request GET \
  --url 'https://api.tokenlane.de/v1/models'
```

## node

```node
import OpenAI from 'openai';

const client = new OpenAI({ baseURL: "https://api.tokenlane.de/v1", apiKey: process.env.TOKENLANE_API_KEY });
const response = await client.models.list({});
console.log(response);
```

## python

```python
import os
import json
from openai import OpenAI

client = OpenAI(base_url="https://api.tokenlane.de/v1", api_key=os.environ["TOKENLANE_API_KEY"])
response = client.models.list(

)
print(response)
```

## vercel

```vercel
// This operation uses HTTP directly.
const response = await fetch("https://api.tokenlane.de/v1/models", {
  method: "GET",
  headers: {
  },
});
console.log(await response.text());
```

## query: input_modalities (optional)

Required comma-separated input modalities (text, image, audio, video, file). A model must support every selected input. Whitespace, empty CSV items and repeated values are normalized; an empty list imposes no input restriction. Unsupported values return 400. Combine with output_modalities=all to include non-text outputs.

### s7: string

{}

## query: output_modalities (optional)

Filter by comma-separated output modalities (text, image, embeddings, audio, video, rerank, decisions, speech, transcription, ocr, pii), or all. Defaults to text, including an empty value.

## query: supported_parameters (optional)

Required comma-separated supported parameters. A model must advertise every named parameter. Whitespace, empty items and repeated values are normalized; an empty list leaves parameters unrestricted. Unknown parameter names match no models.

## query: min_arena_elo (optional)

Minimum Elo of the unambiguous curated primary variant; models without it are excluded

### s24: number

double

{}

## query: max_arena_elo (optional)

Maximum Elo of the unambiguous curated primary variant

## query: arena (optional)

LMArena board for Elo filters; defaults to text; agent IPS is never treated as Elo

## query: category (optional)

Source category for Elo filters; defaults to overall

## query: min_intelligence_index (optional)

Accepted as an opaque query value but not evaluated; reported in x-ignored-params because no Artificial Analysis data is licensed

### s0: string | null

{}

## query: max_intelligence_index (optional)

Accepted as an opaque query value but not evaluated; reported in x-ignored-params

## query: min_coding_index (optional)

Accepted as an opaque query value but not evaluated; reported in x-ignored-params

## query: max_coding_index (optional)

Accepted as an opaque query value but not evaluated; reported in x-ignored-params

## query: min_agentic_index (optional)

Accepted as an opaque query value but not evaluated; reported in x-ignored-params

## query: max_agentic_index (optional)

Accepted as an opaque query value but not evaluated; reported in x-ignored-params

## query: min_tool_success_rate (optional)

Accepted as an opaque query value but not evaluated; reported in x-ignored-params

## query: max_tool_success_rate (optional)

Accepted as an opaque query value but not evaluated; reported in x-ignored-params

## Response 200: application/json

The model catalog, sorted by model ID

```json
{
  "data": [
    {
      "architecture": {
        "input_modalities": [
          "image",
          "text"
        ],
        "modality": "image+text->text",
        "output_modalities": [
          "text"
        ]
      },
      "canonical_slug": "deepseek/deepseek-v4.1-flash",
      "context_length": 131072,
      "created": 0,
      "default_parameters": {
        "temperature": 1,
        "top_p": 0.9
      },
      "description": "Fast general-purpose model for chat, extraction and tool use.",
      "id": "deepseek/deepseek-v4.1-flash",
      "links": {
        "details": "/v1/models"
      },
      "name": "deepseek/deepseek-v4.1-flash",
      "per_request_limits": {
        "completion_tokens": 8192,
        "prompt_tokens": 131072
      },
      "pricing": {
        "audio": "0.00000002",
        "audio_output": "0.00000002",
        "audio_second": "0.00000002",
        "character": "0.00000002",
        "completion": "0.0000008",
        "currency": "EUR",
        "image": "0.00000002",
        "image_output": "0.00000002",
        "image_token": "0.00000002",
        "input_audio_cache": "0.00000002",
        "input_cache_read": "0.00000002",
        "input_cache_write": "0.00000002",
        "input_cache_write_1h": "0.00000002",
        "input_character": "0.000005",
        "internal_reasoning": "0.00000002",
        "page": "0.002",
        "prompt": "0.0000002",
        "request": "0.00000002",
        "web_search": "0.00000002"
      },
      "supported_parameters": [
        "max_tokens",
        "reasoning",
        "response_format",
        "seed",
        "stop",
        "temperature",
        "tool_choice",
        "tools",
        "top_p"
      ],
      "supported_voices": null,
      "top_provider": {
        "context_length": 131072,
        "is_moderated": false,
        "max_completion_tokens": 8192
      }
    }
  ],
  "links": {
    "next": null
  },
  "total_count": 1
}
```

### s25: object

Every model the gateway can serve right now, with capabilities and prices.

{}

- data (required): s338

### s338: array

The models, sorted by ID.

{}

### s163: object

A model with what it can do and what it costs. A model can be served by several providers
(routes); capabilities are those of at least one route, prices those of the cheapest.

{}

- architecture (required): s861

### s861: object

What the model accepts and produces (text, images, ...).

{}

- input_modalities (required): s483

### s483: array

Accepted inputs: `text`, `image`, `video`, `audio`, `file`.

{}

```json
[
  "image",
  "text"
]
```

### s1124: string

Actual input content families. Profile validation rejects output-only modality tags
before publishing a configuration snapshot or this catalog.

"text", "image", "file", "audio", "video"

{}

- modality (required): s484

### s484: string | null

Inputs and outputs in one string (`input+input->output`), or null if unknown.

{}

```json
"image+text->text"
```

- output_modalities (required): s485

### s485: array

What the model can produce: `text`, `image`, `embeddings`, `audio`, `video`,
`rerank` (relevance scores), `transcription` (transcribed text), `speech`, `ocr`,
`pii` or `decisions` (typed structured decisions, served only by `POST /systemone`).

{}

```json
[
  "text"
]
```

- benchmarks (optional): s862

### s862: object

Published benchmark observations and attribution; absent when no source data is available.

{}

- design_arena (required): s878

### s878: array

Required OpenRouter field; empty because no licensed Design Arena data is imported.

{}

### s114: object

OpenRouter's Design Arena entry shape; this gateway publishes no unlicensed entries.

{}

- arena (required): s619

### s619: string

Source arena.

{}

- category (required): s620

### s620: string

Category within the source arena.

{}

- elo (required): s621

### s621: number

Source Elo rating.

double

{}

- rank (required): s622

### s622: integer

Source rank, starting at one.

int32

{"minimum":"0"}

- win_rate (required): s623

### s623: number

Source win percentage.

double

{}

- lmarena (required): s879

### s879: array

Published LMArena variants, with Elo and IPS kept distinct.

{}

### s149: allOf

One published board, category and explicitly named source variant.

{}

allOf: s826

### s826: oneOf

Unchanged score, interval and observation count in the source metric.

{}

oneOf: s486

### s486: object

Human pairwise votes fitted with the source's Bradley-Terry model.

{}

- elo (required): s1125

### s1125: number

Published Elo rating, unchanged from the source.

double

{}

```json
1300.5
```

- elo_lower (required): s1126

### s1126: number

Lower confidence bound for this rating.

double

{}

```json
1290
```

- elo_upper (required): s1127

### s1127: number

Upper confidence bound for this rating.

double

{}

```json
1310
```

- metric (required): s1128

### s1128: string

"bradley_terry"

{}

- votes (required): s1129

### s1129: integer

Human votes counted by the source.

int64

{"minimum":"0"}

```json
12000
```

oneOf: s487

### s487: object

Source IPS observations, with their own unit and confidence interval.

{}

- metric (required): s1130

### s1130: string

"ips"

{}

- observations (required): s1131

### s1131: integer

Agent observations counted by the source, not human votes.

int64

{"minimum":"0"}

```json
1200
```

- score (required): s1132

### s1132: number

Published IPS score; not comparable with Elo ratings.

double

{}

```json
0.75
```

- score_lower (required): s1133

### s1133: number

Lower confidence bound for the IPS score.

double

{}

```json
0.7
```

- score_upper (required): s1134

### s1134: number

Upper confidence bound for the IPS score.

double

{}

```json
0.8
```

- session_count (optional): s1135

### s1135: integer | null

Number of source sessions, absent when unpublished.

int64

{"minimum":"0"}

```json
100
```

allOf: s827

### s827: object

{}

- arena (required): s1185

### s1185: string

Source arena, such as text, vision, webdev or agent.

{}

```json
"text"
```

- as_of (required): s1186

### s1186: string

Publication date of this board, YYYY-MM-DD.

{}

```json
"2026-10-27"
```

- category (required): s1187

### s1187: string

Source category within the arena; overall is the unfiltered board.

{}

```json
"overall"
```

- fetched_at (required): s1188

### s1188: string

UTC timestamp when the complete source snapshot was imported.

{}

```json
"2026-10-27T12:00:00Z"
```

- primary_variant (optional): s1189

### s1189: boolean

Whether this name is the curated first LMArena mapping, also used by the Web catalog.

{}

```json
true
```

- rank (optional): s1190

### s1190: integer | null

Published rank, null when the source does not publish one.

int32

{"minimum":"0"}

```json
3
```

- rank_of (required): s1191

### s1191: integer

Number of source participants in this board, not merely models in this gateway.

int32

{"minimum":"0"}

```json
100
```

- source_name (required): s1192

### s1192: string

Exact source model variant; variants are never silently substituted.

{}

```json
"synthetic-model"
```

- source_row (required): s1193

### s1193: integer

Stable source row index within the imported snapshot.

int32

{"minimum":"0"}

```json
1
```

- sources (required): s880

### s880: array

Attribution for every source in this object.

{}

### s71: object

Required attribution travels with every nonempty model benchmark projection.

{}

- as_of (required): s488

### s488: string

Timestamp of the most recently imported observation in this projection.

{}

```json
"2026-10-27T12:00:00Z"
```

- citation (required): s489

### s489: string

Required citation and license statement.

{}

```json
"Source: LMArena (lmarena.ai), CC BY 4.0"
```

- id (required): s490

### s490: string

Source identifier, lmarena.

{}

```json
"lmarena"
```

- license (required): s491

### s491: string

SPDX license identifier.

{}

```json
"CC-BY-4.0"
```

- license_url (required): s492

### s492: string

License text and attribution requirements.

{}

```json
"https://creativecommons.org/licenses/by/4.0/"
```

- url (required): s493

### s493: string

Public source website.

{}

```json
"https://lmarena.ai"
```

- canonical_slug (required): s863

### s863: string

The model that `id` stands for: equal to `id` for a model; for an alias such as
`eu/fast`, the model the alias currently resolves to, which responses name in `model`.
An entry is an alias exactly when `canonical_slug` differs from `id`.

{}

```json
"deepseek/deepseek-v4.1-flash"
```

- context_length (required): s864

### s864: integer

Longest context (prompt plus answer) in tokens that at least one route accepts. Longer
requests fail with 400 `context_length_exceeded` before any provider is called.

int32

{"minimum":"0"}

```json
131072
```

- created (required): s865

### s865: integer

When the model was added to the catalog, in Unix seconds; 0 if not recorded.

int64

{}

```json
0
```

- default_parameters (required): s866

### s866: object | null

Sampling values a model uses when a request leaves them out. Fields the provider does not
publish are left out.

{}

```json
{
  "temperature": 1,
  "top_p": 0.9
}
```

- frequency_penalty (optional): s1203

### s1203: number | null

Default `frequency_penalty`.

double

{}

- presence_penalty (optional): s1204

### s1204: number | null

Default `presence_penalty`.

double

{}

- repetition_penalty (optional): s1205

### s1205: number | null

Default `repetition_penalty`.

double

{}

- temperature (optional): s1206

### s1206: number | null

Default `temperature`.

double

{}

- top_k (optional): s1207

### s1207: integer | null

Default `top_k`.

int64

{}

- top_p (optional): s1208

### s1208: number | null

Default `top_p`.

double

{}

- description (optional): s867

### s867: string

What the model is good for, in plain text; omitted where none is curated.

{}

```json
"Fast general-purpose model for chat, extraction and tool use."
```

- id (required): s868

### s868: string

The ID to put into `model` of a request, identical to the OpenRouter ID.

{}

```json
"deepseek/deepseek-v4.1-flash"
```

- links (required): s869

### s869: object

Where to fetch details about the model.

{}

```json
{
  "details": "/v1/models"
}
```

- details (required): s887

### s887: string

Path of the catalog that contains this model.

{}

```json
"/v1/models"
```

- name (required): s870

### s870: string

Name for display; currently equal to `id`.

{}

```json
"deepseek/deepseek-v4.1-flash"
```

- per_request_limits (required): s871

### s871: object | null

Token limits per request beyond the context length.

{}

```json
{
  "completion_tokens": 8192,
  "prompt_tokens": 131072
}
```

- completion_tokens (required): s1209

### s1209: integer

Most output tokens per request.

int32

{"minimum":"0"}

```json
8192
```

- prompt_tokens (required): s1210

### s1210: integer

Most input tokens per request.

int32

{"minimum":"0"}

```json
131072
```

- pricing (required): s872

### s872: object

Base prices as decimal strings, with optional ordered whole-request overrides, like
OpenRouter, but in **EUR** instead of USD; hence
the extra field `currency`. Multiply by 1,000,000 for the price per million tokens, the unit
of `provider.max_price.prompt` and `.completion` in requests: `"0.0000002"` is 0.20 EUR per million tokens.
Converted display prices round up to twelve decimal places; filtering and ranking use
their exact values. Positive prices below one pico therefore display as `"0.000000000001"`.

{}

- audio (optional): s406

### s406: string

Price per input audio token. Unset uses the prompt price.

{}

```json
"0.00000002"
```

- audio_output (optional): s407

### s407: string

Price per output audio token. Unset uses the completion price.

{}

```json
"0.00000002"
```

- audio_second (optional): s408

### s408: string

Price per started second of transcribed audio; a gateway extension.

{}

```json
"0.00000002"
```

- character (optional): s409

### s409: string

Price per Unicode scalar value scanned for PII; a gateway extension.

{}

```json
"0.00000002"
```

- completion (required): s1211

### s1211: string

EUR per ordinary output token; also used for reasoning unless priced separately.

{}

```json
"0.0000008"
```

- currency (required): s1212

### s1212: string

Currency of every price and cost in the API. Always `EUR` (not part of OpenRouter's
format, where prices are USD).

{}

```json
"EUR"
```

- image (optional): s410

### s410: string

Price per input image, in addition to any input tokens.

{}

```json
"0.00000002"
```

- image_output (optional): s411

### s411: string

Price per output image, in addition to any output tokens.

{}

```json
"0.00000002"
```

- image_token (optional): s412

### s412: string

Price per uncached input image token. Unset uses the prompt price.

{}

```json
"0.00000002"
```

- input_audio_cache (optional): s413

### s413: string

Price per cached input audio token. Unset uses the audio price.

{}

```json
"0.00000002"
```

- input_cache_read (optional): s414

### s414: string

Price per cached input token read. Unset uses the ordinary prompt price.

{}

```json
"0.00000002"
```

- input_cache_write (optional): s415

### s415: string

Price per token written to cache. Unset uses the ordinary prompt price.

{}

```json
"0.00000002"
```

- input_cache_write_1h (optional): s416

### s416: string

Price per token written to a one-hour cache. Unset uses input_cache_write.

{}

```json
"0.00000002"
```

- input_character (optional): s417

### s417: string

Price per Unicode scalar value submitted for speech synthesis; a gateway extension.

{}

```json
"0.000005"
```

- internal_reasoning (optional): s418

### s418: string

Price per internal reasoning token. Unset uses the completion price.

{}

```json
"0.00000002"
```

- original (optional): s1213

### s1213: allOf

The provider's own prices when they are in another currency and were converted into
EUR with an exchange rate; absent for prices in EUR. Not part of OpenRouter's format.

{}

```json
{
  "completion": "0.00000013",
  "currency": "USD",
  "exchange_rate": {
    "as_of": "2026-10-17T14:00:00Z",
    "markup": "0",
    "per_eur": "1.1645",
    "version": "ecb-2026-10-17"
  },
  "prompt": "0.00000003"
}
```

allOf: s918

### s918: object

Cache, media, request and server-tool prices in this object's currency.

{}

```json
{
  "input_cache_read": "0.00000002",
  "request": "0.001"
}
```

- audio (optional): s406

Price per input audio token. Unset uses the prompt price.

- audio_output (optional): s407

Price per output audio token. Unset uses the completion price.

- audio_second (optional): s408

Price per started second of transcribed audio; a gateway extension.

- character (optional): s409

Price per Unicode scalar value scanned for PII; a gateway extension.

- image (optional): s410

Price per input image, in addition to any input tokens.

- image_output (optional): s411

Price per output image, in addition to any output tokens.

- image_token (optional): s412

Price per uncached input image token. Unset uses the prompt price.

- input_audio_cache (optional): s413

Price per cached input audio token. Unset uses the audio price.

- input_cache_read (optional): s414

Price per cached input token read. Unset uses the ordinary prompt price.

- input_cache_write (optional): s415

Price per token written to cache. Unset uses the ordinary prompt price.

- input_cache_write_1h (optional): s416

Price per token written to a one-hour cache. Unset uses input_cache_write.

- input_character (optional): s417

Price per Unicode scalar value submitted for speech synthesis; a gateway extension.

- internal_reasoning (optional): s418

Price per internal reasoning token. Unset uses the completion price.

- page (optional): s419

### s419: string

Price per processed OCR page; a gateway extension.

{}

```json
"0.002"
```

- request (optional): s420

### s420: string

Fixed charge for one upstream request.

{}

```json
"0.00000002"
```

- web_search (optional): s421

### s421: string

Price per native web search.

{}

```json
"0.00000002"
```

allOf: s919

### s919: object

{}

- completion (required): s1231

### s1231: string

Price per output token in `currency`.

{}

```json
"0.00000013"
```

- currency (required): s1168

### s1168: string

Currency of the provider's price list (ISO 4217).

{}

```json
"USD"
```

- exchange_rate (required): s1232

### s1232: object

The exchange rate the EUR prices are converted with.

{}

```json
{
  "as_of": "2026-10-17T14:00:00Z",
  "markup": "0",
  "per_eur": "1.1645",
  "version": "ecb-2026-10-17"
}
```

- as_of (required): s653

### s653: string

When the rates were published, RFC 3339 in UTC.

date-time

{}

```json
"2026-10-17T14:00:00Z"
```

- markup (required): s654

### s654: string

Markup for exchange-rate risk as a decimal fraction (`"0.02"` is 2 %).

{}

```json
"0"
```

- per_eur (required): s655

### s655: string

Units of the currency one euro buys (ECB convention), as a decimal string.

{}

```json
"1.1645"
```

- version (required): s656

### s656: string

Version of the rate snapshot.

{}

```json
"ecb-2026-10-17"
```

- overrides (optional): s1233

### s1233: array

Ordered conditional price replacements; absent when the base prices always apply.

{}

### s169: allOf

A whole-request conditional price replacement. All present conditions must match;
entries are applied in list order and later matching fields win. Unspecified prices
remain unchanged. Conditions use all actual input tokens, including cache, and the
request's UTC arrival instant. Pre-call routing uses the validated prompt estimate.

{}

```json
{
  "completion": "0.000008",
  "currency": "EUR",
  "min_prompt_tokens": 128000,
  "prompt": "0.000002"
}
```

allOf: s890

### s890: object

Other replacement unit prices, at the same level in the JSON object.

{}

```json
{
  "input_cache_read": "0.00000002",
  "request": "0.001"
}
```

- audio (optional): s406

Price per input audio token. Unset uses the prompt price.

- audio_output (optional): s407

Price per output audio token. Unset uses the completion price.

- audio_second (optional): s408

Price per started second of transcribed audio; a gateway extension.

- character (optional): s409

Price per Unicode scalar value scanned for PII; a gateway extension.

- image (optional): s410

Price per input image, in addition to any input tokens.

- image_output (optional): s411

Price per output image, in addition to any output tokens.

- image_token (optional): s412

Price per uncached input image token. Unset uses the prompt price.

- input_audio_cache (optional): s413

Price per cached input audio token. Unset uses the audio price.

- input_cache_read (optional): s414

Price per cached input token read. Unset uses the ordinary prompt price.

- input_cache_write (optional): s415

Price per token written to cache. Unset uses the ordinary prompt price.

- input_cache_write_1h (optional): s416

Price per token written to a one-hour cache. Unset uses input_cache_write.

- input_character (optional): s417

Price per Unicode scalar value submitted for speech synthesis; a gateway extension.

- internal_reasoning (optional): s418

Price per internal reasoning token. Unset uses the completion price.

- page (optional): s419

Price per processed OCR page; a gateway extension.

- request (optional): s420

Fixed charge for one upstream request.

- web_search (optional): s421

Price per native web search.

allOf: s891

### s891: object

{}

- completion (optional): s1218

### s1218: string

Replacement price per ordinary output token.

{}

```json
"0.000008"
```

- currency (required): s1219

### s1219: string

Currency of this entry's prices, the same as its base price set.

{}

```json
"EUR"
```

- min_prompt_tokens (optional): s1220

### s1220: integer

Strict lower bound on total prompt tokens: 128000 matches 128001 and above.

int64

{"minimum":"0"}

```json
128000
```

- prompt (optional): s1221

### s1221: string

Replacement price per ordinary input token.

{}

```json
"0.000002"
```

- utc_days (optional): s1222

### s1222: array

Weekdays at the request instant: monday through sunday. Absent means all.

{}

```json
[
  "monday",
  "tuesday"
]
```

- utc_end (optional): s1223

### s1223: integer

Exclusive UTC end as HHMM. End before start wraps midnight; equal times span 24h.

int32

{"minimum":"0"}

```json
630
```

- utc_start (optional): s1224

### s1224: integer

Inclusive UTC start as HHMM. Requires utc_end; for example 1630 means 16:30 UTC.

int32

{"minimum":"0"}

```json
1630
```

- prompt (required): s1234

### s1234: string

Price per input token in `currency`.

{}

```json
"0.00000003"
```

- overrides (optional): s1214

### s1214: array

Ordered conditional price replacements; absent when the base prices always apply.

{}

- page (optional): s419

Price per processed OCR page; a gateway extension.

- prompt (required): s1215

### s1215: string

EUR per input token (prompt, including images and tools as tokens).

{}

```json
"0.0000002"
```

- request (optional): s420

Fixed charge for one upstream request.

- web_search (optional): s421

Price per native web search.

- reasoning (optional): s873

### s873: object

How the model reasons and which efforts it serves; omitted for models without reasoning.

{}

```json
{
  "default_effort": "low",
  "default_enabled": true,
  "mandatory": false,
  "supported_efforts": [
    "xhigh",
    "medium",
    "low"
  ]
}
```

- default_effort (optional): s892

### s892: string | null

Effort used when a request sets neither an effort, a budget nor `enabled`, the same on
every route; `none` means reasoning is off unless requested. Absent when the
provider's own default applies.

{}

```json
"low"
```

- default_enabled (optional): s893

### s893: boolean

Whether the model reasons when the request does not say; absent if unknown.

{}

```json
true
```

- mandatory (required): s894

### s894: boolean

True when no route can switch reasoning off. Effort `none` or `enabled: false` then
runs the lowest effort and removes the reasoning from the response, as
`reasoning.exclude` does; the reasoning tokens are still billed.

{}

```json
false
```

- supported_efforts (optional): s895

### s895: array | null

Efforts at least one route serves as such, highest first, without `none`. Null when a
route passes any effort to its provider unchanged.

{}

```json
[
  "xhigh",
  "medium",
  "low"
]
```

### s1225: string

Curated reasoning levels published by the catalog. The route projection selects these
from the provider dialect's finite effort domain and excludes the disabled level `none`.

"max", "xhigh", "high", "medium", "low", "minimal"

{}

- routes (optional): s874

### s874: array

Route-level compliance and data handling; guarantees apply to the named route only.

{}

### s171: object

Verified privacy metadata of one available model/provider route.

{}

- compliance (required): s896

### s896: string

Jurisdiction guarantee, strongest first: sovereign_eu (EU company and EU-owned
infrastructure), eu_operated (EU contracting party and operator, EU data on
non-EU-owned infrastructure),
eu_data_residency (EU data, non-EU company or parent), global (unrestricted).

{}

```json
"sovereign_eu"
```

- context_length (optional): s897

### s897: integer | null

Maximum combined context tokens on this route, absent if unknown.

int32

{"minimum":"0"}

- data_policy (required): s898

### s898: object

Provider default or complete route override, with conservative unknown values.

{}

- datacenters (required): s883

### s883: array

Documented processing countries in ISO alpha-2 form; empty when unknown.

{}

```json
[
  "FR"
]
```

- retention (required): s884

### s884: string

Retention conditions including metadata; unknown if not established.

{}

```json
"No content retention; billing metadata retained for 12 months."
```

- trains_on_data (required): s885

### s885: boolean

True also when no training guarantee is established; false means no training.

{}

```json
false
```

- zdr (required): s886

### s886: boolean

True only when zero retention of contents is explicitly established.

{}

```json
true
```

- max_completion_tokens (optional): s899

### s899: integer | null

Maximum output tokens on this route, absent if unknown.

int32

{"minimum":"0"}

- pricing (required): s900

### s900: anyOf

Published EUR prices including cache and other components; null when not representable.

{}

anyOf: s168

### s168: allOf

Base prices as decimal strings, with optional ordered whole-request overrides, like
OpenRouter, but in **EUR** instead of USD; hence
the extra field `currency`. Multiply by 1,000,000 for the price per million tokens, the unit
of `provider.max_price.prompt` and `.completion` in requests: `"0.0000002"` is 0.20 EUR per million tokens.
Converted display prices round up to twelve decimal places; filtering and ranking use
their exact values. Positive prices below one pico therefore display as `"0.000000000001"`.

{}

```json
{
  "completion": "0.0000008",
  "currency": "EUR",
  "prompt": "0.0000002"
}
```

allOf: s888

### s888: object

Cache, media, request and server-tool prices in this object's currency.

{}

```json
{
  "input_cache_read": "0.00000002",
  "request": "0.001"
}
```

- audio (optional): s406

Price per input audio token. Unset uses the prompt price.

- audio_output (optional): s407

Price per output audio token. Unset uses the completion price.

- audio_second (optional): s408

Price per started second of transcribed audio; a gateway extension.

- character (optional): s409

Price per Unicode scalar value scanned for PII; a gateway extension.

- image (optional): s410

Price per input image, in addition to any input tokens.

- image_output (optional): s411

Price per output image, in addition to any output tokens.

- image_token (optional): s412

Price per uncached input image token. Unset uses the prompt price.

- input_audio_cache (optional): s413

Price per cached input audio token. Unset uses the audio price.

- input_cache_read (optional): s414

Price per cached input token read. Unset uses the ordinary prompt price.

- input_cache_write (optional): s415

Price per token written to cache. Unset uses the ordinary prompt price.

- input_cache_write_1h (optional): s416

Price per token written to a one-hour cache. Unset uses input_cache_write.

- input_character (optional): s417

Price per Unicode scalar value submitted for speech synthesis; a gateway extension.

- internal_reasoning (optional): s418

Price per internal reasoning token. Unset uses the completion price.

- page (optional): s419

Price per processed OCR page; a gateway extension.

- request (optional): s420

Fixed charge for one upstream request.

- web_search (optional): s421

Price per native web search.

allOf: s889

### s889: object

{}

- completion (required): s1211

EUR per ordinary output token; also used for reasoning unless priced separately.

- currency (required): s1212

Currency of every price and cost in the API. Always `EUR` (not part of OpenRouter's
format, where prices are USD).

- original (optional): s1216

### s1216: allOf

The provider's own prices when they are in another currency and were converted into
EUR with an exchange rate; absent for prices in EUR. Not part of OpenRouter's format.

{}

```json
{
  "completion": "0.00000013",
  "currency": "USD",
  "exchange_rate": {
    "as_of": "2026-10-17T14:00:00Z",
    "markup": "0",
    "per_eur": "1.1645",
    "version": "ecb-2026-10-17"
  },
  "prompt": "0.00000003"
}
```

allOf: s918

allOf: s919

- overrides (optional): s1217

### s1217: array

Ordered conditional price replacements; absent when the base prices always apply.

{}

- prompt (required): s1215

EUR per input token (prompt, including images and tools as tokens).

anyOf: s588

### s588: null

{}

- provider_name (required): s901

### s901: string

Provider ID; the same provider may serve a model with several tariff variants.

{}

```json
"ovh"
```

- quantization (optional): s902

### s902: string | null

Verified deployment precision, absent when unknown.

{}

```json
"fp8"
```

- region (optional): s903

### s903: string

Actual route processing region as an assigned uppercase ISO 3166-1 alpha-2 country code;
omitted when unknown, never inferred from data_policy.datacenters.

{"minLength":"2","maxLength":"2","pattern":"\"^[A-Z]{2}$\""}

```json
"FR"
```

- supported_parameters (required): s904

### s904: array

Request parameters supported by this route, independently of the model-wide union.

{}

- tag (required): s905

### s905: string

Endpoint tag to name this route in `provider.only`, `provider.order` or
`provider.ignore`: the provider ID for the standard route, `<provider>/<variant>`
for a tariff variant, like OpenRouter's endpoint `tag`.

{}

```json
"grunden/cache"
```

- variant (optional): s906

### s906: string | null

Tariff variant of the provider (e.g. `cache`: a separately priced tariff with cheap
cache reads); omitted for the standard route. Variants serve only requests naming
their `tag` in `provider.only` or `provider.order`, never by fallback or sorting, and
are excluded from model-wide `pricing` unless the model has no standard route.

{}

```json
"cache"
```

- supported_parameters (required): s875

### s875: array

Request parameters at least one route supports. Others are dropped before the request
reaches the provider, or rule the route out with `provider.require_parameters: true`.
`parallel_tool_calls` means at least one route can enable parallel tool calls;
serial tool routes may still accept `parallel_tool_calls: false` without listing it.

{}

```json
[
  "max_tokens",
  "reasoning",
  "response_format",
  "seed",
  "stop",
  "temperature",
  "tool_choice",
  "tools",
  "top_p"
]
```

- supported_voices (required): s876

### s876: array | null

Voices of speech models; null for models without speech output.

{}

- top_provider (required): s877

### s877: object

Context, output limit and moderation of the route tried first.

{}

```json
{
  "context_length": 131072,
  "is_moderated": false,
  "max_completion_tokens": 8192
}
```

- context_length (optional): s995

### s995: integer | null

Longest context of this route in tokens, or null if unknown.

int32

{"minimum":"0"}

```json
131072
```

- is_moderated (required): s996

### s996: boolean

Whether the provider filters content on this route; filtered answers end with
`finish_reason: "content_filter"`.

{}

```json
false
```

- max_completion_tokens (optional): s997

### s997: integer | null

Most output tokens this route generates per request, or null if unknown; the remaining
context window can limit the answer further.

int32

{"minimum":"0"}

```json
8192
```

- links (required): s339

### s339: object

Pagination; the catalog always fits on one page, so `next` is null.

{}

```json
{
  "next": null
}
```

- next (required): s920

### s920: string | null

URL of the next page; null, because the catalog is returned in full.

{}

- total_count (required): s340

### s340: integer

Number of models in `data`.

{"minimum":"0"}

```json
1
```

## Response 400: application/json

Invalid, duplicated or conflicting benchmark query parameters

```json
{
  "error": {
    "code": 400,
    "message": "Use a finite numeric Elo threshold",
    "metadata": {
      "docs_url": "/docs#description/invalid-parameter",
      "issues": [
        {
          "code": "invalid_value",
          "message": "Use a finite numeric Elo threshold",
          "path": "query.min_arena_elo"
        }
      ],
      "remedy_hint": "Check the named benchmark query parameters. Elo filters use the curated primary variant; agent IPS is not Elo.",
      "request_id": "req_01k7tm9qz3f5c8x2v6b4n0m7hd",
      "retryable": false,
      "type": "invalid_parameter"
    }
  }
}
```

### s3: object

Every response with an HTTP status of 400 or higher has this shape, OpenRouter's error
format. Everything the gateway adds is inside `error.metadata`, so OpenRouter clients see
nothing unfamiliar. See "Errors" in the introduction for all error types.

{}

```json
{
  "error": {
    "code": 400,
    "message": "deepseek/deepseek-v4.1-flsh is not a valid model ID",
    "metadata": {
      "docs_url": "/docs#description/model-not-found",
      "issues": [
        {
          "code": "unknown_model",
          "message": "deepseek/deepseek-v4.1-flsh is not a valid model ID",
          "path": "$.model"
        }
      ],
      "remedy_hint": "Did you mean deepseek/deepseek-v4.1-flash?",
      "request_id": "req_01k7tm9qz3f5c8x2v6b4n0m7hd",
      "retryable": false,
      "type": "model_not_found"
    }
  },
  "user_id": "org_01k7tm2c4e6g8j0k2m4p6r8t0v"
}
```

- error (required): s238

### s238: object

What went wrong.

{}

```json
{
  "code": 400,
  "message": "temperature must be at most 2, got 5 (and 1 more)",
  "metadata": {
    "docs_url": "/docs#description/invalid-parameter",
    "issues": [
      {
        "code": "out_of_range",
        "message": "temperature must be at most 2, got 5",
        "path": "$.temperature"
      },
      {
        "code": "invalid_value",
        "message": "messages[1].role must be one of system, developer, user, assistant, tool, got \"bot\"",
        "path": "$.messages[1].role"
      }
    ],
    "request_id": "req_01k7tm9qz3f5c8x2v6b4n0m7hd",
    "retryable": false,
    "type": "invalid_parameter"
  }
}
```

- code (required): s633

### s633: integer

The HTTP status of the response, as a number.

int32

{"minimum":"0"}

```json
400
```

- message (required): s634

### s634: string

What went wrong, in English. With several problems: the first one and how many more
follow; all of them are in `metadata.issues`. The text may change; match on
`metadata.type` instead.

{}

```json
"temperature must be at most 2, got 5"
```

- metadata (optional): s635

### s635: object

Details for programs: error type, request ID, every problem, a suggested fix and, for
provider errors, which provider failed. Always present on gateway errors.

{}

```json
{
  "docs_url": "/docs#description/rate-limited",
  "provider_error": {
    "category": "rate_limited",
    "provider_name": "deepinfra",
    "retry_after": 2,
    "retryable": true,
    "status": 429
  },
  "provider_name": "deepinfra",
  "remedy_hint": "Retry after the Retry-After interval, or allow fallbacks to other providers (provider.allow_fallbacks).",
  "request_id": "req_01k7tm9qz3f5c8x2v6b4n0m7hd",
  "retry_after": 2,
  "retryable": true,
  "type": "rate_limited"
}
```

- available_providers (optional): s639

### s639: array

Providers that serve the model (when `provider.only` excluded all of them).

{}

```json
[
  "deepinfra",
  "open-inference"
]
```

- docs_url (required): s640

### s640: string

Link to the section of this error type in the documentation.

{}

```json
"/docs#description/model-not-found"
```

- failed_routing_step (optional): s641

### s641: string

The routing step that left no route (routing errors only).

{}

```json
"Filter by Max Price"
```

- issues (optional): s642

### s642: array

Every problem found in the request, so all can be fixed at once (at most 20); for
routing errors one entry per excluded route, for `no_provider_available` one per failed
attempt.

{}

### s123: object

One problem in the request, one reason a route was excluded, or one failed attempt.

{}

```json
{
  "code": "invalid_url",
  "message": "messages[2].content[0].image_url.url must be an http(s) URL or a base64 data URL",
  "path": "$.messages[2].content[0].image_url.url"
}
```

- code (required): s636

### s636: string

Kind of problem, stable for programs.

"required", "invalid_type", "invalid_value", "out_of_range", "empty", "invalid_url", "invalid_base64", "unsupported_media_type", "media_type_mismatch", "media_too_large", "too_many_media", "unsupported_media_transport", "media_fetch_failed", "unknown_tool", "unknown_model", "not_available", "not_allowed", "route_excluded", "bridge_failed", "provider_failed"

{}

```json
"out_of_range"
```

- message (required): s637

### s637: string

The problem in English, naming the field and, if short, the value sent.

{}

```json
"temperature must be at most 2, got 5"
```

- path (required): s638

### s638: string

Where the problem is: JSONPath of the field in the request body, for example
`$.messages[2].content[0].image_url.url`, or `query.<name>` for a query parameter.

{}

```json
"$.temperature"
```

- provider_error (optional): s643

### s643: oneOf

Safe provider diagnostics, identical for HTTP errors and terminal stream events.

{}

oneOf: s1158

### s1158: object

Provider diagnostics without provider-authored text. HTTP and terminal SSE errors use
this same structure; arbitrary upstream messages and metadata are discarded.

{}

```json
{
  "category": "rate_limited",
  "provider_name": "deepinfra",
  "retry_after": 2,
  "retryable": true,
  "status": 429
}
```

- category (required): s1299

### s1299: string

Gateway classification from HTTP status and recognized dialect identifiers.

"rate_limited", "overloaded", "invalid_request", "context_length", "content_filtered", "authentication", "billing", "timeout", "unavailable", "unknown"

{}

```json
"rate_limited"
```

- code (optional): s1300

### s1300: string

Exact identifier from the configured dialect's fixed list; arbitrary values vanish.

{}

```json
"rate_limit_exceeded"
```

- provider_name (required): s1301

### s1301: string

Configured provider name, never a name supplied by the upstream body.

{}

```json
"deepinfra"
```

- retry_after (optional): s1302

### s1302: integer

Bounded seconds from the upstream Retry-After header, when present and numeric.

int64

{"minimum":"0"}

```json
2
```

- retryable (required): s1303

### s1303: boolean

Whether the classified failure can be transient. Unknown failures are not promised retryable.

{}

- status (optional): s1304

### s1304: integer

Actual HTTP status received from the upstream, including 200 for native errors.
Absent when no HTTP status is available.

int32

{"minimum":"0"}

```json
429
```

oneOf: s588

- provider_name (optional): s644

### s644: string

The provider that failed, as in the response field `provider` (provider errors only).

{}

```json
"deepinfra"
```

- raw (optional): s645

### s645: string

Reserved OpenRouter compatibility field. The gateway never fills it: provider
free text is deliberately withheld. Use provider_error instead.

{}

```json
""
```

- remedy_hint (optional): s646

### s646: string

A concrete fix, for example the closest model ID to a mistyped one.

{}

```json
"Did you mean deepseek/deepseek-v4.1-flash?"
```

- request_id (required): s647

### s647: string

ID of the request, the same as the `x-request-id` response header. Quote it when
contacting support.

{}

```json
"req_01k7tm9qz3f5c8x2v6b4n0m7hd"
```

- requested_providers (optional): s648

### s648: array

Providers named in `provider.only` (when it excluded all routes).

{}

```json
[
  "scaleway"
]
```

- retry_after (optional): s649

### s649: integer

Seconds to wait before retrying; also sent as the `Retry-After` header (429 and 503).

int64

{"minimum":"0"}

```json
2
```

- retryable (required): s650

### s650: boolean

Whether sending the identical request again can succeed (for example after a timeout
or rate limit), as opposed to errors that need a changed request.

{}

- routing_funnel (optional): s651

### s651: array

How routing narrowed down the providers: each step with the number of routes left
(routing errors only, same format as OpenRouter).

{}

### s200: object

One routing step, as in OpenRouter's `routing_funnel`.

{}

```json
{
  "endpoint_count": 0,
  "reason": "provider.max_price",
  "step": "Filter by Max Price"
}
```

- endpoint_count (required): s979

### s979: integer

Number of routes (providers of the model) left after the step.

{"minimum":"0"}

```json
2
```

- reason (optional): s980

### s980: string

The request field the step filtered on.

{}

```json
"provider.only"
```

- step (required): s981

### s981: string

Name of the step, as at OpenRouter.

{}

```json
"Initial Endpoints"
```

- type (required): s652

### s652: string

Stable error type to match on in code; see "Errors" in the introduction. Messages may
change, the type does not.

"invalid_request", "invalid_parameter", "unsupported_parameter", "context_length_exceeded", "authentication", "insufficient_credits", "workspace_budget_exceeded", "permission_denied", "moderation", "model_not_found", "no_endpoint", "not_found", "method_not_allowed", "timeout", "payload_too_large", "media_fetch_failed", "rate_limited", "provider_error", "idempotency_conflict", "idempotency_in_progress", "idempotency_unavailable", "provider_invalid_response", "no_provider_available", "byok_invalid_credential", "byok_rate_limited", "byok_unavailable", "feature_unavailable", "internal"

{}

```json
"invalid_parameter"
```

- user_id (optional): s239

### s239: string | null

Organization of the API key, if the key was recognized.

{}

```json
"org_01k7tm2c4e6g8j0k2m4p6r8t0v"
```
