Anthropic Claude Model Catalog
Werker Health supports every generally available Anthropic Claude model. Use the native model string (e.g., claude-sonnet-4-5) exactly as published by Anthropic once your SDK is pointed at https://api.werker.health/v1.
✅ Anthropic’s Claude 4.5 family introduces frontier-grade intelligence while maintaining strict safety guardrails. Consult the Claude model overview for live availability details.
Latest Frontier Models
| Model | Description |
|---|---|
claude-sonnet-4-5 | Smartest Claude model for complex agents, coding, and enterprise reasoning. |
claude-haiku-4-5 | Fastest Claude option with near-frontier intelligence for latency-sensitive tasks. |
claude-opus-4-1 | Exceptional model for specialized, high-stakes reasoning workloads. |
Claude 3.5 Lineup
| Model | Highlights |
|---|---|
claude-3-5-sonnet-latest | Flagship reasoning model with strong coding, multilingual, and long-context performance. |
claude-3-5-haiku-latest | High-speed, cost-efficient model tuned for near real-time interactions. |
claude-3-5-opus-preview | Advanced reasoning for complex agent workflows (preview availability). |
Claude 3 Series
| Model | Highlights |
|---|---|
claude-3-sonnet-20240229 | Production-grade balance of quality, latency, and cost. |
claude-3-opus-20240229 | Deep reasoning model for enterprise-grade analysis. |
claude-3-haiku-20240307 | Lightweight, fast responses for chatbots and workflow automations. |
Legacy and Specialized Models
| Model | Highlights |
|---|---|
claude-instant-1.2 | Legacy low-latency model for backward compatibility. |
claude-3-sonnet-20240229-json | Optimized for strict JSON compliance in function-style outputs. |
Using Claude Models with Werker Health
- Model strings: Use Anthropic’s published Claude IDs directly—no additional prefix required.
- Safety systems: Werker preserves Anthropic’s constitutional AI guardrails and augments them with Werker PHI policies.
- Streaming support: Claude streaming tokens are forwarded in real time with immutable logging.
- Version pinning: Anthropic maintains
-latestaliases; pin exact dates for deterministic behavior in regulated workflows.
Source: Anthropic Claude Models
Sample Usage
python
from openai import OpenAI
client = OpenAI(
base_url="https://api.werker.health/v1",
api_key="WERKER_HEALTH_API_KEY",
)
plan = client.responses.create(
model="claude-sonnet-4-5",
input=[
{"role": "system", "content": "Draft clinician-friendly visit summaries."},
{"role": "user", "content": "Summarize the telehealth visit that just concluded."},
],
)
print(plan.output_text)ts
import OpenAI from "openai";
const client = new OpenAI({
baseURL: "https://api.werker.health/v1",
apiKey: "WERKER_HEALTH_API_KEY",
});
const plan = await client.responses.create({
model: "claude-sonnet-4-5",
input: [
{ role: "system", content: "Draft clinician-friendly visit summaries." },
{ role: "user", content: "Summarize the telehealth visit that just concluded." },
],
});
console.log(plan.output_text);