Skip to main content

google/gemini-4-carbon/chat

Google's Gemini 4 Carbon chat endpoint runs its strongest reasoning model on text and images, with function calling, JSON output and a 1,048,576-token shared context.

Input $2Output $10per 1 million tokensCached input $0.10

Workbench

Send a request

Quickstart

Examples use the selected endpoint’s public parameter definitions. Replace the API key and example media, then send the request from your server.

SDKs & integrations
POSThttps://api.spicyapi.ai/v1/chat/completions

API authentication

Keep the API key in the server environment variable SPICY_API_KEY. Use Bearer authentication and Content-Type: application/json for JSON requests.

Authorization: Bearer $SPICY_API_KEY
curl --fail-with-body -sN https://api.spicyapi.ai/v1/chat/completions \
  -H "Authorization: Bearer $SPICY_API_KEY" \
  -H "Content-Type: application/json" \
  -H "Idempotency-Key: $(uuidgen)" \
  -d '{
  "model": "google/gemini-4-carbon/chat",
  "messages": [
    {
      "role": "user",
      "content": "Summarise the following release notes into three bullet points."
    }
  ],
  "stream": true
}' \
  | jq --unbuffered -Rrj 'select(startswith("data: ")) | .[6:] | select(. != "[DONE]") | fromjson | if .error then "\n[error] \(.error.message)\n" else (.choices[0].delta.content? // empty) end'

Generated from this model’s input schema — the same one the API validates every request against.

Keep one Idempotency-Key for each logical request. After a network interruption, do not automatically resend with a new key; retain request_id to investigate the accepted request. Idempotency does not replay an earlier chat stream.

Input parameters

Place the model parameters directly in the JSON request body alongside model. This endpoint does not use an input wrapper.

messagesarray<object>Required

The Carbon conversation, oldest first. User messages can combine text parts with image_url parts.

Default
Not specified
Constraints
minItems:1
maxItems:256
response_formatobjectOptional

Asks Carbon for a JSON object, or for JSON that follows a json_schema you provide. Validate the result before your application relies on it.

Default
Not specified
tool_choicestring | objectOptional

Sets Carbon's next move: auto, none, required, or one named function.

Default
Not specified
parallel_tool_callsbooleanOptional

On by default, so Carbon can request several functions in one turn. Set it to false to receive one call at a time.

Default
true
toolsarray<object>Optional

Up to 128 function definitions Carbon can request. Your application validates the arguments, runs the function and returns the result in a tool message.

Default
Not specified
Constraints
maxItems:128
top_pnumberOptional

Accepted from 0.01 to 1 so OpenAI-style clients keep working. Google's documentation says Carbon ignores this setting as well.

Default
Not specified
Constraints
minimum:0.01
maximum:1
temperaturenumberOptional

Accepted from 0 to 2 so OpenAI-style clients keep working. Google's documentation says Carbon ignores this setting, so it will not change the reply.

Default
Not specified
Constraints
minimum:0
maximum:2
stopstring | arrayOptional

One to four stop strings of up to 32 characters each. Carbon's reply ends before the first match, and the match is not returned.

Default
Not specified
Constraints
minLength:1
minItems:1
maxItems:4
reasoning_effortstringOptional

Carbon thinking depth: low, medium or high, with medium as the default. Thinking cannot be turned off, and high uses the most tokens.

Default
"medium"
Allowed values
"low""medium""high"
max_tokensintegerOptional

Carbon reply budget from 1 to 65536, default 4096. Reasoning is counted here: a 400-token cap on a long prompt used 397 tokens on reasoning and returned no text.

Default
4096
Constraints
minimum:1
maximum:65536

Read the result

With stream set to false, the request returns the completed chat response directly. Read choices[].message; no task polling is needed. Set stream to true when your client supports Server-Sent Events.

Completed response example
{
  "id": "chatcmpl_example",
  "object": "chat.completion",
  "created": 1788652800,
  "model": "google/gemini-4-carbon/chat",
  "choices": [
    {
      "index": 0,
      "message": {
        "role": "assistant",
        "content": "Here is a concise summary of the release notes."
      },
      "finish_reason": "stop"
    }
  ],
  "usage": {
    "prompt_tokens": 32,
    "completion_tokens": 14,
    "total_tokens": 46
  }
}

The response example is illustrative. usage reports token counts; token counts and charges depend on your actual request.

SDKs & integrations

Use SpicyAPI in your app, terminal, AI assistant or visual workflow. Choose a setup guide below.

Use it in SillyTavern and roleplay apps

In SillyTavern, RisuAI, Agnai or any app with an OpenAI-compatible setting, paste the base URL below and your API key, then pick google/gemini-4-carbon/chat as the model. Browser calls are allowed, and sliders this model does not take are ignored instead of rejected.

Base URL: https://api.spicyapi.ai/v1
Model: google/gemini-4-carbon/chat

Use it in ComfyUI

Install the SpicyAPI nodes from ComfyUI Manager or with the command below, then pick google/gemini-4-carbon/chat in the SpicyAPI Chat node to use this model inside a ComfyUI graph.

cd ComfyUI/custom_nodes
git clone https://github.com/spicyapi-ai/comfyui-spicyapi