Which endpoint and authentication are required?

Send a POST request to `https://api.evolink.ai/v1/chat/completions` and include `Authorization: Bearer ` plus `Content-Type: application/json` using an API key from your EvoLink dashboard.

Gemini 3.1 Pro CustomTools API

Use the request model ID gemini-3.1-pro-customtools through EvoLink's OpenAI-compatible /v1/chat/completions endpoint. On EvoLink, this is a separately exposed Gemini 3.1 Pro family route for tool-heavy workflows.

Model Type:

✓Gemini 3.1 Pro CustomTools

Price:

$1.600 - 3.200(~ 115.2 - 230.4 credits) per 1M input tokens; $9.600 - 14.400(~ 691.2 - 1036.8 credits) per 1M output tokens

$0.161 - 0.322(~ 11.6 - 23.2 credits) per 1M cache read tokens

Google Search grounding charged separately per query.

Highest stability with guaranteed 99.9% uptime. Recommended for production environments.

Use the same API endpoint for all versions. Only the model parameter differs.

PRICING

PLAN	CONTEXT WINDOW	MAX OUTPUT	INPUT	OUTPUT	CACHE READ
Gemini 3.1 Pro CustomTools	1,050,000	65,536	≤200.0K$1.600-20% (115.2 Credits) >200.0K$3.200-20% (230.4 Credits)	≤200.0K$9.600-20% (691.2 Credits) >200.0K$14.400-20% (1,036.8 Credits)	≤200.0K$0.161-19% (11.6 Credits) >200.0K$0.322-19% (23.2 Credits)
Web Search Tool Server-side web search capability					$0.011/search (0.81 Credits)

Pricing Note: Prices show both USD and Credits. Units default to / 1M tokens unless noted separately.

Cache Hit: Price applies to cached prompt tokens.

Gemini 3.1 Pro CustomTools API

Gemini 3.1 Pro CustomTools is a separate request route and model ID on EvoLink for tool-heavy Gemini 3.1 Pro workflows. Use it when your requests rely on tools arrays, structured outputs, and multi-step agent loops, and keep the base Gemini 3.1 Pro route for broader general reasoning.

Model route

Request model ID

gemini-3.1-pro-preview-customtools

Best use cases for Gemini 3.1 Pro CustomTools API

Tool-Centric Request Flows

Use this route when your workload is organized around a tools array, structured outputs, and repeated back-and-forth between the model and external systems. It fits agent loops, tool-driven extraction, and orchestrated API workflows.

AI Agent & Automation Workflows

Build agent workflows that call external APIs, databases, or internal services through one gateway. On EvoLink, this model ID gives you a distinct tool-focused route inside the Gemini 3.1 Pro family instead of mixing that traffic into the base Pro route.

1M Context + Structured Outputs

Use the 1M context window for longer agent sessions, larger tool payloads, and structured extraction jobs. This route keeps the pricing and model ID separate from the base Pro page, which makes planning and routing clearer.

Why Use Gemini 3.1 Pro CustomTools on EvoLink

EvoLink exposes Gemini 3.1 Pro CustomTools as its own request route and model ID, so teams can keep one gateway and one auth pattern while separating tool-heavy Gemini traffic from the base Gemini 3.1 Pro route.

Keep One Gateway for Agent Routing

You can add gemini-3.1-pro-customtools to an existing OpenAI-style stack without creating a separate Google-specific integration path for one workflow branch.

Separate General Reasoning from Tool-Oriented Traffic

Using a separate model ID makes it easier to keep agent traffic, structured output jobs, and general Gemini 3.1 Pro usage on distinct routes with clearer cost and behavior controls.

Production-Grade Token Accounting

Usage includes prompt_tokens, completion_tokens, total_tokens, plus completion_tokens_details.reasoning_tokens for tracking tool-planning overhead separately from final output.

How to Call Gemini 3.1 Pro CustomTools

Use the OpenAI SDK format with your existing tool definitions and the gemini-3.1-pro-customtools model string.

Step 1 - Set the Model

Use model: "gemini-3.1-pro-customtools" in the request body. Your existing tool/function definitions work without changes.

Step 2 - Define Tools and Send Messages

Include a tools array with your function definitions and a messages array with role/content pairs. This route is intended for workloads where tool selection, structured outputs, and multi-turn tool use are central to the request flow.

Step 3 - Handle Tool Calls and Results

Process tool_calls in the response, execute the functions, and send results back. Use the live pricing table on this page to confirm the route cost before moving the workflow into production.

Gemini 3.1 Pro CustomTools API Features and Limits

Core capabilities optimized for tool-heavy workloads

Tools

Enhanced Function Calling

Designed for routes where function calling, tools arrays, and structured outputs are part of the normal request shape.

Agent

Multi-Step Orchestration

Handles chained tool invocations where outputs feed into subsequent calls with context retention.

Output

Structured Outputs

Reliable JSON schema conformance for structured data extraction and API response formatting.

Context

1M Context Window

1,048,576 input tokens with tiered pricing above 200K for long agent sessions and large tool contexts.

Analytics

Reasoning Tokens

completion_tokens_details includes reasoning_tokens to track tool-planning overhead separately.

Compatibility

OpenAI SDK Compatible

Use the standard /v1/chat/completions interface with tools array — no custom integration needed.

Gemini 3.1 Pro CustomTools API FAQs

Everything you need to know about the product and billing.

On EvoLink, Gemini 3.1 Pro CustomTools is a separately exposed request route and model ID inside the Gemini 3.1 Pro family. It is intended for tool-heavy workflows, not just as another label on the base Gemini 3.1 Pro page.

The practical difference on EvoLink is route selection and workload fit. Use Gemini 3.1 Pro for broader general-purpose multimodal reasoning, and use gemini-3.1-pro-customtools when you want a separate request route for tool-oriented traffic, structured outputs, and agent-style orchestration.

Use "gemini-3.1-pro-customtools" as the model value for the Gemini 3.1 Pro CustomTools API on EvoLink. This is the recommended model string for the OpenAI-compatible chat completions endpoint.

Send a POST request to `api.evolink.ai/v1/chat/completions` and include `Authorization: Bearer <token>` plus `Content-Type: application/json` using an API key from your EvoLink dashboard.

Yes. CustomTools is fully compatible with the OpenAI tools array format. Your existing function definitions, parameter schemas, and tool_choice settings work without modification.

Input token limit is 1,048,576 (1M context) and max output tokens are 65,536. Input pricing has tiered rates: standard pricing up to 200K tokens, and a higher tier above 200K tokens for long-context usage.

Yes. This route is intended for workflows where the model may need to read tool definitions, choose tools, receive tool results, and continue across multiple turns within the same request loop.

Continue with Gemini family pages and integration guides

Where Gemini 3.1 Pro CustomTools fits in the Gemini family

Treat CustomTools as a separate tool-oriented request route inside your Gemini 3.1 Pro family options on EvoLink. Use it when function calling, structured outputs, and agent orchestration are the main workload pattern; keep the base Gemini 3.1 Pro route for broader general-purpose reasoning traffic.

Group family-model links and integration content in one place so the page stays focused and the next step is clearer.

Gemini family routes

Gemini 3.1 Pro

The base Gemini 3.1 Pro model for general multimodal reasoning without tool-use optimization.

Gemini 3.1 Flash Lite

Low-cost Flash Lite route for high-throughput translation, extraction, and classification workloads.

Integration and migration

Gemini migration guide

Review a real Gemini migration path that keeps the OpenAI SDK request shape.

Gemini CLI gateway setup

Run Gemini CLI, Codex CLI, and Claude Code through one gateway.