Together AI Launches Kimi K3 for Long-Context Agentic Workflows

Together AI announced as the Day 0 launch partner for Moonshot AI's Kimi K3, which is now available via its high-throughput inference service. Designed for long-running, tool-intensive agentic workflows, the model offers developers a production-ready deployment path.

Confirmed

Core parameters and features: Kimi K3 has a total parameter count of 2.8T, supports a context window of up to 1M, and features native vision capabilities. Its primary use cases are explicitly geared toward coding agents and long-horizon task workflows.

Service and deployment: Together AI highlights that the model is hosted on US-based infrastructure with a strict zero-data retention policy. An API has been released for developer integration.

Why it matters

This launch provides AI agent developers handling ultra-long contexts and complex tool calls with a high-throughput, privacy-focused option, further accelerating the deployment of heavy-duty agent applications in production environments.

2026-07-28 ~ 2026-07-28 · 7 related posts

• [source] Together AI brings Moonshot’s Kimi K3 online with 1M context and agent tools — togethercompute · 2026-07-28
• Kimi K3 launches with 2.8T parameters, 1M context and $0.30 input pricing — togethercompute · 2026-07-28
• Kimi K3 launches on Together AI with Day 0 access for coding agents — togethercompute · 2026-07-28
• Kimi K3 goes live on Together AI for long-running agentic workflows — togethercompute · 2026-07-28
• [source] Together AI says Kimi K3 targets long tool-heavy agent workflows with 1M context — togethercompute · 2026-07-28
• [source] Together AI pitches US-hosted Kimi K3 API with zero data retention — togethercompute · 2026-07-28

1 near-duplicate retellings: togethercompute