> ## Documentation Index
> Fetch the complete documentation index at: https://docs.bytespike.ai/llms.txt
> Use this file to discover all available pages before exploring further.

# GET /models

> List the models your key can call, with capability tags and live pricing pointers.

The catalog endpoint. Returns the models in scope for your API key —
the same set that `model` field accepts on
[`/v1/messages`](../text/claude-messages),
[`/v1/chat/completions`](../text/openai-chat-completions), and
[`/v1beta/models/{model}:generateContent`](../text/gemini-generate-content).
Calling it is free and never debits credits.

## When to use

* **Verify key validity** at boot — a 200 here proves the key is alive and in-scope
* **Discover capability tags** before sending an image / video / audio block
* **Build a model picker** in your own UI without hard-coding the catalog
* **Detect retired slugs** — anything not in the list returns 400 `unsupported_model` on `/messages` / `/chat/completions`

## Request

```bash theme={null}
curl https://llm.bytespike.ai/v1/models \
  -H "Authorization: Bearer $BYTESPIKE_API_KEY"
```

`x-api-key: $BYTESPIKE_API_KEY` also works. No body.

### Query parameters

| Param        | Type   | Notes                                                                                                                                  |
| ------------ | ------ | -------------------------------------------------------------------------------------------------------------------------------------- |
| `capability` | string | Filter by capability tag, e.g. `?capability=vision` or `?capability=tool_use`. Multiple OK: `?capability=vision&capability=streaming`. |
| `family`     | string | Filter by model family: `anthropic` / `openai` / `google` / `deepseek` / `moonshot` / `zhipu` / `minimax` / `bytedance`.               |

## Response

```json theme={null}
{
  "object": "list",
  "data": [
    {
      "id": "claude-sonnet-4-6",
      "object": "model",
      "owned_by": "anthropic",
      "family": "claude",
      "capabilities": ["text", "vision", "tool_use", "streaming", "cache_control", "thinking"],
      "context_window": 200000,
      "max_output_tokens": 8192,
      "input_modalities": ["text", "image"],
      "output_modalities": ["text"],
      "pricing_url": "https://bytespike.ai/pricing#claude-sonnet-4-6",
      "created": 1714003200
    },
    {
      "id": "gpt-5-4",
      "object": "model",
      "owned_by": "openai",
      "family": "gpt",
      "capabilities": ["text", "vision", "tool_use", "streaming", "json_schema", "logprobs"],
      "context_window": 256000,
      "max_output_tokens": 16384,
      "input_modalities": ["text", "image"],
      "output_modalities": ["text"],
      "pricing_url": "https://bytespike.ai/pricing#gpt-5-4",
      "created": 1716393600
    }
    // …
  ]
}
```

### Response fields

| Field                      | Type      | Notes                                                                                                |
| -------------------------- | --------- | ---------------------------------------------------------------------------------------------------- |
| `data[].id`                | string    | Model slug. Use this as the `model` value on inference endpoints.                                    |
| `data[].owned_by`          | string    | Model originator (`anthropic`, `openai`, `google`, etc.).                                            |
| `data[].family`            | string    | Coarser grouping for UI (`claude`, `gpt`, `gemini`, `deepseek`, `kimi`, `glm`, `minimax`, `doubao`). |
| `data[].capabilities`      | string\[] | Feature tags (see [Capability tags](#capability-tags)).                                              |
| `data[].context_window`    | integer   | Max input context in tokens.                                                                         |
| `data[].max_output_tokens` | integer   | Hard cap on a single generation.                                                                     |
| `data[].input_modalities`  | string\[] | `text`, `image`, `audio`, `video`.                                                                   |
| `data[].output_modalities` | string\[] | Same union; text-only models return `["text"]`.                                                      |
| `data[].pricing_url`       | string    | Deep link to the per-model pricing row.                                                              |
| `data[].created`           | integer   | Unix timestamp when the model entered the catalog.                                                   |

### Live pricing

The `pricing_url` deep-links to the row in [bytespike.ai/pricing](https://bytespike.ai/pricing). We don't ship per-token rates in this response on purpose — the rate card is refreshed nightly from the gateway and the docs page reflects whatever the gateway will actually bill you at, so reading it client-side stays canonical.

## Capability tags

| Tag             | Meaning                                                                                                            |
| --------------- | ------------------------------------------------------------------------------------------------------------------ |
| `text`          | Accepts text input + emits text output. Present on every model.                                                    |
| `vision`        | Accepts image content blocks (`image` on `/messages`, `image_url` on `/chat/completions`, `inlineData` on Gemini). |
| `audio`         | Accepts audio input parts.                                                                                         |
| `video`         | Accepts video input parts.                                                                                         |
| `tool_use`      | Honors `tools` / `tool_choice`.                                                                                    |
| `streaming`     | Supports `stream: true` SSE.                                                                                       |
| `cache_control` | Supports Anthropic-style prompt caching.                                                                           |
| `thinking`      | Emits `thinking` content blocks (Opus / Sonnet 4.x).                                                               |
| `json_schema`   | Supports OpenAI `response_format: {"type": "json_schema"}`.                                                        |
| `logprobs`      | Supports OpenAI `logprobs: true`.                                                                                  |
| `image_gen`     | Image generation endpoint (`/v1/images/generations`).                                                              |
| `video_gen`     | Video generation endpoint (`/v1/videos/generations`).                                                              |

## Errors

| Status | `error.type`           | Trigger                                                                                     |
| ------ | ---------------------- | ------------------------------------------------------------------------------------------- |
| 401    | `authentication_error` | Missing / revoked key.                                                                      |
| 403    | `permission_error`     | Key valid but no models granted (rare — usually means the key was created in a paused org). |

`GET /v1/models` is free; no `402` path. Rate limits exist but are very loose — the only realistic 429 is a key-rotation hammer storm.

## Example — boot-time validation

```python theme={null}
import os
import requests

r = requests.get(
    "https://llm.bytespike.ai/v1/models",
    headers={"Authorization": f"Bearer {os.environ['BYTESPIKE_API_KEY']}"},
    timeout=10,
)
if r.status_code == 401:
    raise SystemExit("BYTESPIKE_API_KEY missing or revoked")
catalog = {m["id"]: m for m in r.json()["data"]}
assert "vision" in catalog["claude-sonnet-4-6"]["capabilities"]
```

## Example — capability filter

```bash theme={null}
curl 'https://llm.bytespike.ai/v1/models?capability=vision&capability=tool_use' \
  -H "Authorization: Bearer $BYTESPIKE_API_KEY"
```

Returns only models that accept images AND can call tools — useful when wiring an agent that has to OCR a screenshot and then act on it.