# StreamLake

> Kuaishou's AI cloud, home of the KAT-Coder agentic coding models.

Canonical: https://www.subconscious.dev/providers/streamlake · By The Subconscious Team · Updated October 11, 2026

- Founded: brand launched around 2024 (Kuaishou dates to 2011)
- Example models: KAT-Coder-Pro V2.5, KAT-Coder-Air V2.5
- Website: https://www.streamlake.ai

## Overview

StreamLake is the AI cloud brand of Kuaishou, the Chinese short-video company behind the Kling video models. It sells model-as-a-service inference and bare-metal compute to internet businesses, drawing on the infrastructure Kuaishou built to serve video at massive scale. Its developer site offers APIs, SDKs and integration guides aimed at taking teams from testing to production, and it also hosts third-party open models like DeepSeek V4 Flash 0731.

The headline model is KAT-Coder-Pro V2.5, a proprietary agentic coding model from Kuaishou's KwaiKAT team. StreamLake says large-scale agentic reinforcement learning trained it to handle repository-level work: reading an issue, finding changes across files, running tests and fixing its own errors over long runs. Developers can pay per token, at $0.74 in and $2.96 out for Pro V2.5 or $0.15 in and $0.60 out for Air V2.5 with 256K context, or buy a KwaiKAT Coding Plan, and both come with OpenAI-protocol endpoints plus a Claude-protocol proxy that drops into Claude Code or OpenClaw. Earlier KAT-Coder versions spread through resellers like 302.AI and OpenRouter.

## Upsides

- Direct access to Kuaishou's in-house coding model, with a Claude Code-compatible endpoint.
- Backed by a large company with hyperscale video infrastructure.

## Core use cases

- Low-cost agentic coding through a subscription plan.
- Chinese internet businesses that want domestic MaaS and bare-metal capacity.

## Downsides

- Much of the documentation still leads with China, though the English product page now lists USD prices.
- Data residency in China rules it out for many US and EU enterprise buyers.

## At a glance

| Attribute | Value |
|---|---|
| Model access | Proprietary coding models, plus hosted open models |
| Flagship models | KAT-Coder-Pro V2.5, KAT-Coder-Air V2.5 |
| Speed | - |
| Price | $0.74 in, $2.96 out (Pro V2.5), or KwaiKAT Coding Plan |
| Customization | - |
| Deployment | MaaS API, bare metal |
| Long context | 256K |

## FAQ

### What is StreamLake?

StreamLake is the AI cloud brand of Kuaishou, the Chinese short-video company behind the Kling video models. It sells model-as-a-service inference and bare-metal compute to internet businesses, drawing on the infrastructure Kuaishou built to serve video at massive scale. Its developer site offers APIs, SDKs and integration guides aimed at taking teams from testing to production, and it also hosts third-party open models like DeepSeek V4 Flash 0731.

### What is StreamLake best for?

Low-cost agentic coding through a subscription plan; Chinese internet businesses that want domestic MaaS and bare-metal capacity.

### How much does StreamLake cost?

StreamLake pricing at a glance: $0.74 in, $2.96 out (Pro V2.5), or KwaiKAT Coding Plan. Rates change often, so check StreamLake's pricing page before committing.

### How much context does StreamLake support?

StreamLake's long-context support: 256K.

### What are the downsides of StreamLake?

Much of the documentation still leads with China, though the English product page now lists USD prices; Data residency in China rules it out for many US and EU enterprise buyers.

### What are the best alternatives to StreamLake?

Common alternatives include Subconscious, OpenAI, Anthropic, Google Vertex AI, Amazon Bedrock. Each has a head-to-head comparison with StreamLake on this site.

## Comparisons

- [Subconscious vs StreamLake](https://www.subconscious.dev/compare/subconscious-vs-streamlake.md)
- [OpenAI vs StreamLake](https://www.subconscious.dev/compare/openai-vs-streamlake.md)
- [Anthropic vs StreamLake](https://www.subconscious.dev/compare/anthropic-vs-streamlake.md)
- [Google Vertex AI vs StreamLake](https://www.subconscious.dev/compare/google-vertex-vs-streamlake.md)
- [Amazon Bedrock vs StreamLake](https://www.subconscious.dev/compare/aws-bedrock-vs-streamlake.md)
- [Together AI vs StreamLake](https://www.subconscious.dev/compare/together-ai-vs-streamlake.md)
- [Fireworks AI vs StreamLake](https://www.subconscious.dev/compare/fireworks-vs-streamlake.md)
- [Baseten vs StreamLake](https://www.subconscious.dev/compare/baseten-vs-streamlake.md)
- [Groq vs StreamLake](https://www.subconscious.dev/compare/groq-vs-streamlake.md)
- [Cerebras vs StreamLake](https://www.subconscious.dev/compare/cerebras-vs-streamlake.md)
- [DeepInfra vs StreamLake](https://www.subconscious.dev/compare/deepinfra-vs-streamlake.md)
- [Hugging Face Inference Providers vs StreamLake](https://www.subconscious.dev/compare/hugging-face-vs-streamlake.md)
- [Modal vs StreamLake](https://www.subconscious.dev/compare/modal-vs-streamlake.md)
- [Cloudflare Workers AI vs StreamLake](https://www.subconscious.dev/compare/cloudflare-workers-ai-vs-streamlake.md)
- [xAI vs StreamLake](https://www.subconscious.dev/compare/xai-vs-streamlake.md)
- [Mistral AI vs StreamLake](https://www.subconscious.dev/compare/mistral-ai-vs-streamlake.md)
- [DeepSeek vs StreamLake](https://www.subconscious.dev/compare/deepseek-vs-streamlake.md)
- [Moonshot AI vs StreamLake](https://www.subconscious.dev/compare/moonshot-ai-vs-streamlake.md)
- [Z.ai vs StreamLake](https://www.subconscious.dev/compare/z-ai-vs-streamlake.md)
- [Alibaba Cloud vs StreamLake](https://www.subconscious.dev/compare/alibaba-cloud-vs-streamlake.md)
- [Meta vs StreamLake](https://www.subconscious.dev/compare/meta-vs-streamlake.md)
- [Cohere vs StreamLake](https://www.subconscious.dev/compare/cohere-vs-streamlake.md)
- [SambaNova vs StreamLake](https://www.subconscious.dev/compare/sambanova-vs-streamlake.md)
- [Nebius vs StreamLake](https://www.subconscious.dev/compare/nebius-vs-streamlake.md)
- [Crusoe vs StreamLake](https://www.subconscious.dev/compare/crusoe-vs-streamlake.md)
- [fal vs StreamLake](https://www.subconscious.dev/compare/fal-vs-streamlake.md)
- [Novita AI vs StreamLake](https://www.subconscious.dev/compare/novita-ai-vs-streamlake.md)
- [Venice vs StreamLake](https://www.subconscious.dev/compare/venice-vs-streamlake.md)
- [Parasail vs StreamLake](https://www.subconscious.dev/compare/parasail-vs-streamlake.md)
- [Inference.net vs StreamLake](https://www.subconscious.dev/compare/inference-net-vs-streamlake.md)
- [GMI Cloud vs StreamLake](https://www.subconscious.dev/compare/gmi-cloud-vs-streamlake.md)
- [Thinking Machines vs StreamLake](https://www.subconscious.dev/compare/thinking-machines-vs-streamlake.md)
- [Sail Research vs StreamLake](https://www.subconscious.dev/compare/sail-research-vs-streamlake.md)
- [Morph vs StreamLake](https://www.subconscious.dev/compare/morph-vs-streamlake.md)
- [Relace vs StreamLake](https://www.subconscious.dev/compare/relace-vs-streamlake.md)
- [TypeSafe AI vs StreamLake](https://www.subconscious.dev/compare/typesafe-ai-vs-streamlake.md)
- [StepFun vs StreamLake](https://www.subconscious.dev/compare/stepfun-vs-streamlake.md)
- [Runware vs StreamLake](https://www.subconscious.dev/compare/runware-vs-streamlake.md)
- [StreamLake vs Wafer](https://www.subconscious.dev/compare/streamlake-vs-wafer.md)
- [StreamLake vs RunInfra](https://www.subconscious.dev/compare/streamlake-vs-runinfra.md)
- [StreamLake vs Particle.AI](https://www.subconscious.dev/compare/streamlake-vs-particle-ai.md)
- [StreamLake vs Luminal](https://www.subconscious.dev/compare/streamlake-vs-luminal.md)
- [StreamLake vs Infron](https://www.subconscious.dev/compare/streamlake-vs-infron.md)
- [StreamLake vs Inception](https://www.subconscious.dev/compare/streamlake-vs-inception.md)
- [StreamLake vs Dolphin](https://www.subconscious.dev/compare/streamlake-vs-dolphin.md)

## Sources

- [StreamLake](https://www.streamlake.ai/)
- [KAT-Coder on StreamLake](https://www.streamlake.ai/product/kat-coder)
- [KwaiKAT Coding Plan](https://www.streamlake.ai/marketing/coding-plan)
- [KAT-Coder on 302.AI](https://doc-en.302.ai/366936071e0)

Pricing and model lineups change often; figures are a snapshot.