One developer experience for agents, tools, streaming, structured output, and provider-aware routing—without locking your application to one model vendor.
import { generateText } from "@zhivex-ai/sdk"
;import { createOpenAI } from "@zhivex-ai/openai"
;const model = createOpenAI()("gpt-6-astra"
);const result = await generateText({ model, prompt: "Ship something remarkable."
});console.log(result.text);Built for a multi-provider world
Provider APIs move fast. Zhivex gives your product a stable application layer while keeping advanced provider features within reach.
Portable coreAgents · tools · streams · stateUse the ecosystem your team already knows. Both SDKs are designed around portable AI foundations and deliberate provider-specific control.
@zhivex-ai/sdk
A unified, ESM-first toolkit for generation, agents, durable workflows, artifacts, observability, provider routing, and production controls.
bun add @zhivex-ai/sdkzhivex-ai-sdk
Build orchestrated AI systems with agents, handoffs, approvals, durable run state, tools, traces, and portable provider foundations.
pip install zhivex-ai-sdkNew multimodal inputs, more resilient agents, and a richer React chat experience. Follow each SDK’s release status before upgrading.
Understand text, images, audio, and video in one request. Available in @zhivex-ai/qwen 0.13.0 for TypeScript, with streaming, hybrid reasoning, callable tools, and hosted web search through Responses.
HTTP Chat Completions and Responses return text only. Structured output uses prompted generation with local schema validation. Python support is available in 0.25.0.
bun add @zhivex-ai/sdk @zhivex-ai/qwenThe latest patch adds Qwen3.8-Omni-Flash support; the 1.21 series also expands agent recovery and native provider APIs.
Version 0.25.0 adds multimodal provider updates and new native APIs. The Python distribution remains Beta.
Opt-in cursor replay and reconnects, attachment uploads with progress and retries, optional Markdown and virtualized messages, and draft recovery after failed sends. Connect the UI to your server-side runtime.
Features vary by provider and API. Catalog entries and release availability do not imply live certification for every model.
Adopt one layer at a time. The shared contract stays small while operational capabilities remain available when your system needs them.
Keep messages, tools, streaming events, usage, and structured output consistent as models change.
Build tool-using agents with sessions, handoffs, approvals, durable state, replay, and governed control-plane primitives.
Add safety policies, evaluation gates, bounded artifacts, OpenTelemetry, model catalogs, and provider-aware routing.
Start portable, then reach provider-specific capabilities without leaking them through your whole app.
Use text, tools, structured output, embeddings, audio, files, and realtime features where supported.
Keep credentials, tools, databases, and policy on trusted server runtimes—not inside browser clients.
Pick your language, copy the command, and follow the focused quickstart for your runtime.
# Install the SDK and a provider
bun add @zhivex-ai/sdk @zhivex-ai/openai
import { generateText } from "@zhivex-ai/sdk";
import { createOpenAI } from "@zhivex-ai/openai";
const openai = createOpenAI();
const result = await generateText({
model: openai("gpt-6-astra"),
prompt: "What should we build next?"
});
console.log(result.text);# Install the SDK
pip install zhivex-ai-sdk
import asyncio
from zhivex_ai import Agent, create_openai, run_agent
async def main():
provider = create_openai()
agent = Agent(
name="builder",
instructions="Help ship useful AI products.",
model=provider("gpt-6-astra"),
)
result = await run_agent(agent=agent, prompt="What should we build next?")
print(result.text)
asyncio.run(main())Begin with generation or the agent runtime and install only the provider adapters your app needs.
Add shared SQL state, approval queues, artifacts, privacy-aware telemetry, and regression gates before real users arrive.
Use the shared contract for portable behavior and native namespaces only for deliberate provider features.