Google Gemini Model Catalog

Werker Health exposes every generally available Gemini model. Use the native Gemini model string (e.g., gemini-2.5-pro) exactly as documented by Google after pointing the SDK at https://api.werker.health/v1.

✅ Gemini model families, capabilities, and limits are summarized below. Refer to Google’s Gemini Models documentation for the authoritative list and updates.

Featured Models

Model	Positioning	Highlights
gemini-2.5-pro	Our most advanced model	Long-context “thinking” model built for complex reasoning across code, math, and STEM domains.
gemini-2.5-flash	Fast and intelligent	Best price-performance for large-scale, low-latency agentic workloads.
gemini-2.5-flash-lite	Ultra fast	Cost-efficient, high-throughput variant optimized for speed-sensitive scenarios.

Additional Gemini 2.5 Variants

Model	Highlights
gemini-2.5-pro-preview-tts	Preview text-to-speech model delivering natural audio outputs (8k input, 16k output tokens).
gemini-2.5-flash-preview-09-2025	Latest flash enhancements released September 2025.

Gemini 2.0 Family

Model	Highlights
gemini-2.0-flash	1M token input, 8k output; structured outputs, Google Maps grounding, and code execution supported.
gemini-2.0-flash-lite	Cost-optimized 1M token context for large batch inference.
gemini-2.0-flash-preview-image-generation	Multimodal model returning text and images; preview availability varies by region.
gemini-2.0-flash-live-001	Real-time audio/text interactions; preview slated for deprecation December 9, 2025.

Previous Generation Models

Model	Highlights
gemini-1.5-pro-latest	Long-context reasoning with multimodal inputs.
gemini-1.5-flash-latest	Fast, cost-effective deployments requiring multimodal awareness.

Using Gemini Models with Werker Health

• Model strings: Use Google’s published Gemini IDs directly—no additional prefix required.
• Audit trails: Werker logs the exact provider model string, enabling cost attribution and compliance reporting per provider.
• Capability flags: Use Werker metadata to track which features (code execution, function calling, grounding) you enable per deployment.
• Environment promotion: Evaluate new preview or experimental Gemini releases in staging before rolling to production.

Source: Google Gemini API Models

Sample Usage

Python

from openai import OpenAI

client = OpenAI(
    base_url="https://api.werker.health/v1",
    api_key="WERKER_HEALTH_API_KEY",
)

response = client.responses.create(
    model="gemini-2.5-pro",
    input=[
        {"role": "system", "content": "Summarize clinical documents in plain language."},
        {"role": "user", "content": "Summarize the attached cardiology consult note."},
    ],
)

print(response.output_text)

TypeScript

import OpenAI from "openai";

const client = new OpenAI({
  baseURL: "https://api.werker.health/v1",
  apiKey: "WERKER_HEALTH_API_KEY",
});

const response = await client.responses.create({
  model: "gemini-2.5-pro",
  input: [
    { role: "system", content: "Summarize clinical documents in plain language." },
    { role: "user", content: "Summarize the attached cardiology consult note." },
  ],
});

console.log(response.output_text);