{"schema_version":"onlylabs.public_analysis.v1","url":"https://onlylabs.fyi/analysis/streamlake","json_url":"https://onlylabs.fyi/analysis/streamlake/analysis.json","evidence_json_url":"https://onlylabs.fyi/analysis/streamlake/evidence.json","generated_at":"2026-06-27T22:22:39.547Z","analysis":{"org_slug":"streamlake","url":"https://onlylabs.fyi/analysis/streamlake","json_url":"https://onlylabs.fyi/analysis/streamlake/analysis.json","evidence_json_url":"https://onlylabs.fyi/analysis/streamlake/evidence.json","dossier_url":"https://onlylabs.fyi/labs/streamlake","org":{"slug":"streamlake","name":"StreamLake (Kuaishou)","category":"neocloud","category_label":"Neocloud","homepage_url":"https://www.streamlake.com"},"title":"StreamLake (Kuaishou) analysis","summary":"Kuaishou's StreamLake is executing a dual-pronged open-source strategy: a video-native multimodal foundation model line (Keye-VL-2.0) aimed at long-video understanding and agentic capabilities, and an AI coding product suite (KAT-Coder, CodeFlicker, Vanchin) targeting the software engineering tool market. Both tracks are anchored in Apache 2.0 releases, rigorous public benchmarking against frontier models (GPT-5,…","markdown":"## Thesis\n\nKuaishou's StreamLake is executing a dual-pronged open-source strategy: a video-native multimodal foundation model line (Keye-VL-2.0) aimed at long-video understanding and agentic capabilities, and an AI coding product suite (KAT-Coder, CodeFlicker, Vanchin) targeting the software engineering tool market. Both tracks are anchored in Apache 2.0 releases, rigorous public benchmarking against frontier models (GPT-5, Gemini, Claude, Qwen), and a conservative scaling posture that prioritizes cost efficiency over infrastructure buildout [W5, W6, W1, W2].\n\n## Signal desks\n\n### Hiring\nNo cited evidence in this pack.\n\n### Forks\n- **kwaipilot/experiments** — fork of **SWE-bench/experiments**, the official SWE-bench leaderboard submission repository for predictions, execution logs, trajectories, and evaluation results. This fork, dated September 2025, indicates internal SWE-bench evaluation infrastructure work that directly precedes the KAT-Coder product launch in October 2025 [P4, E4, W5].\n\n### Releases\n- **Keye-VL-2.0-30B-A3B** (May 2026): 30B total parameters, MoE architecture with ~3B active parameters. First multimodal model to integrate DeepSeek Sparse Attention (DSA). 256K context window for hour-long video temporal reasoning. Debuts Code Interpreter, Tool Use, and Search agent capabilities in the Keye line. Apache 2.0, weights on Hugging Face and ModelScope, public GitHub repo, demo available [W1, W2, W3, W4].\n- **SWE-Compass** (December 2025): Apache 2.0 evaluation framework covering 8 software engineering task types, 8 programming scenarios, and 10 programming languages, with 2,000 instances sourced from real GitHub PRs. arXiv paper published. Hugging Face dataset available [P1, E1].\n- **KAT-Coder-Agent** (September 2025): Agent repo created with minimal public detail; 1 star, 0 forks [P2, E2].\n- **KAT-Coder** (September 2025): Repository created with HTML language tag; sparse public detail; 1 star, 0 forks [P3, E3].\n- **AI Coding Product Matrix** (October 2025): CodeFlicker (intelligent development tool), multiple self-developed KAT-Coder models, and Vanchin (万擎) large model platform. KAT-Coder-Air V1 offered free to all users [W5](https://36kr.com/newsflashes/3521399817051270).\n\n### Talking\n- **Keye-VL-2.0 launch narrative**: Framed as the \"first to bring DSA into multimodal understanding,\" delivering \"near-lossless end-to-end temporal reasoning on hour-long video,\" explicitly benchmarked as beating Gemini-2.5-Pro and Gemini 3 Flash on TimeLens video metrics [W2, W3]. External coverage highlights beating Qwen3-VL-235B on LongVideoBench at 74.1 with one-eighth the active parameters, positioning the model as a \"credible drop-in for long-video agent stacks built on closed APIs\" [W1](https://ai-tldr.dev/releases/kwai-keye-vl-2-0-30b-a3b/).\n- **Coding product launch narrative**: 36Kr coverage frames the AI coding release as a \"工具+模型+平台\" (tools + models + platform) three-in-one matrix, with the specific claim that KAT-Coder-Pro V1 surpasses GPT-5 and Claude Sonnet 4 on SWE-bench Verified at 73.4% [W5](https://36kr.com/newsflashes/3521399817051270).\n- **StreamLake video cloud positioning**: Conservative scaling rhetoric — \"will not massively invest in infrastructure\" or \"burn money expanding teams,\" instead growing \"like a snowball\" with healthy margins. AI focus explicitly on video AI across the creation-to-distribution pipeline, validated against real large-scale data [W6](https://www.36kr.com/p/1866758221189888).\n\n## Shipping\n\nEvidence of five distinct shipped artifacts across two product lines:\n\n1. **Keye-VL-2.0-30B-A3B**: Model weights on Hugging Face (`Kwai-Keye/Keye-VL-2.0-30B-A3B`) and ModelScope, GitHub repository (`Kwai-Keye/Keye`), public demo, Apache 2.0 license. Shipped May 2026 with Code Interpreter, Tool Use, and Search agent features [W1, W2, W3, W4].\n2. **SWE-Compass**: Evaluation framework on GitHub (`kwaipilot/SWE-Compass`), Hugging Face dataset, arXiv paper (2511.05459), Apache 2.0. Shipped December 2025 [P1, E1].\n3. **AI Coding product matrix**: CodeFlicker, KAT-Coder models (Pro V1 at 73.4% SWE-bench Verified, Air V1 free), Vanchin platform. Launched October 2025 [W5](https://36kr.com/newsflashes/3521399817051270).\n4. **KAT-Coder-Agent**: Minimal public release, September 2025 [P2, E2].\n5. **KAT-Coder**: Minimal public release, September 2025 [P3, E3].\n\n## Research themes\n\n- **Multi-dimensional SWE evaluation**: SWE-Compass extends beyond Python-centric SWE-bench with 10 languages, 8 task types, and 8 programming scenarios, sourced from real GitHub pull requests. Explicitly targets gaps in \"narrow task categories, Python-centric bias, and insufficient alignment with real-world development workflows\" [P1](https://github.com/kwaipilot/SWE-Compass).\n- **Sparse attention for multimodal models**: Keye-VL-2.0 is claimed as the first multimodal model to integrate DeepSeek Sparse Attention (DSA), enabling a 256K context window for \"almost lossless reasoning\" on long video [W2, W3, W4].\n- **Long-video temporal reasoning**: Benchmark results on LongVideoBench (74.1) and QVHighlights-TimeLens (70.1 mIoU) demonstrate competitive or superior performance against larger models (Qwen3-VL-235B, Gemini-2.5-Pro) for hour-long video understanding [W1, W2].\n- **Agent-augmented multimodal models**: Keye-VL-2.0 introduces Code Interpreter, Tool Use, and Search capabilities, evolving the model \"from passive observer to active agent\" [W2, W3].\n- **AI coding agents and evaluation**: SWE-bench experiments fork for evaluation pipeline management, KAT-Coder agent development, and SWE-Compass benchmark construction form a coherent coding-agent research thread [P1, P2, P3, P4, W5].\n\n## Hiring & scaling\n\nNo cited evidence in this pack. Public positioning from StreamLake leadership states the video cloud unit \"will not massively invest in infrastructure\" or \"burn money expanding teams,\" preferring a \"snowball\" growth model [W6](https://www.36kr.com/p/1866758221189888). However, no specific job listings, team composition data, or location-based hiring signals are present in the supplied evidence.\n\n## Category implications\n\n- **Video AI open-weight competition**: Keye-VL-2.0's Apache 2.0 release, combined with benchmark wins against larger closed and open models (Gemini-2.5-Pro, Qwen3-VL-235B), positions Kuaishou as a serious open-weight entrant in long-video understanding. The permissive license makes it a \"credible drop-in for long-video agent stacks built on closed APIs\" [W1](https://ai-tldr.dev/releases/kwai-keye-vl-2-0-30b-a3b/), which could pressure commercial video understanding API pricing [W1, W2, W4].\n- **AI coding tools market entry**: KAT-Coder-Pro V1 beating GPT-5 and Claude Sonnet 4 on SWE-bench Verified (73.4%) combined with a free tier (Air V1) signals an aggressive product entry against incumbent coding assistants. The \"工具+模型+平台\" three-in-one approach suggests a platform ecosystem play rather than a point-product [W5](https://36kr.com/newsflashes/3521399817051270).\n- **Evaluation infrastructure as moat**: SWE-Compass and the SWE-bench experiments fork indicate significant investment in proprietary evaluation methodology that extends beyond community benchmarks. Building a 10-language, 8-task-type benchmark from real GitHub PRs creates evaluation capabilities that directly feed model development and product benchmarking [P1, P4].\n- **Platform GTM strategy**: The pairing of open-weight model releases (Keye-VL-2.0, SWE-Compass) with commercial platform products (Vanchin, CodeFlicker) mirrors the cloud-to-enterprise funnel seen in other neocloud plays. However, StreamLake's stated conservative scaling posture — \"no large-scale infrastructure investment\" and \"no burning cash on team expansion\" — suggests a capital-efficient approach that may limit near-term capacity growth [W5, W6].\n- **Research-to-product pipeline**: The September 2025 SWE-bench experiments fork [E4](https://github.com/kwaipilot/experiments) followed by KAT-Coder repos in mid-September [E2, E3] and the full product matrix launch in October 2025 [W5](https://36kr.com/newsflashes/3521399817051270) demonstrates a compressed research-to-product cycle of roughly one month. SWE-Compass in December 2025 [E1](https://github.com/kwaipilot/SWE-Compass) and Keye-VL-2.0 in May 2026 [W1, W2] show sustained release cadence across both tracks.\n- **Thin evidence areas**: No hiring data, no infrastructure or compute procurement signals, no revenue or customer metrics, and no evidence of safety, alignment, or policy work. The agent capabilities claimed for Keye-VL-2.0 (Code Interpreter, Tool Use, Search) are described but not independently benchmarked in the cited sources [W2, W3]. KAT-Coder-Agent and KAT-Coder repos remain at 1 star each with minimal documentation, suggesting pre-release or limited public launch status [P2, P3].\n\n## Traction highlights\n\n- **Keye-VL-2.0**: 74.1 on LongVideoBench, beating Qwen3-VL-235B (one-eighth the active parameters); 70.1 mIoU on QVHighlights-TimeLens, beating Gemini-2.5-Pro and Gemini 3 Flash [W1, W2]\n- **KAT-Coder-Pro V1**: 73.4% on SWE-bench Verified, surpassing GPT-5 and Claude Sonnet 4 [W5](https://36kr.com/newsflashes/3521399817051270)\n- **SWE-Compass**: 18 GitHub stars, 2 forks, Hugging Face dataset published [P1](https://github.com/kwaipilot/SWE-Compass)\n- **KAT-Coder-Agent**: 1 star, 0 forks — pre-traction [P2](https://github.com/kwaipilot/KAT-Coder-Agent)\n- **KAT-Coder**: 1 star, 0 forks — pre-traction [P3](https://github.com/kwaipilot/KAT-Coder)\n- **Community attention**: Keye-VL-2.0 covered by AI/TLDR, AI Tech Deep Dives, CSDN, and AI铺子 across English and Chinese tech media [W1, W2, W3, W4]; AI coding launch covered by 36Kr [W5](https://36kr.com/newsflashes/3521399817051270)\n\n## Sources\n\n- P1: kwaipilot/SWE-Compass repository — evaluation framework, Apache 2.0, 18 stars\n- P2: kwaipilot/KAT-Coder-Agent repository — agent repo, 1 star\n- P3: kwaipilot/KAT-Coder repository — coding model repo, 1 star\n- P4: kwaipilot/experiments — fork of SWE-bench/experiments\n- E1: kwaipilot/SWE-Compass repo creation event (2025-12-03)\n- E2: kwaipilot/KAT-Coder-Agent repo creation event (2025-09-16)\n- E3: kwaipilot/KAT-Coder repo creation event (2025-09-16)\n- E4: kwaipilot/experiments fork event from SWE-bench/experiments (2025-09-10)\n- W1: AI/TLDR — Keye-VL-2.0-30B-A3B release coverage\n- W2: AI Tech Deep Dives — Keye-VL-2.0-30B-A3B technical analysis\n- W3: CSDN blog — Kuaishou Keye-VL-2.0 launch announcement\n- W4: AI铺子 — Keye-VL-2.0-30B-A3B summary\n- W5: 36Kr — StreamLake AI Coding product matrix launch\n- W6: 36Kr — StreamLake video cloud brand launch and strategy","generated_at":"2026-06-27T19:15:04.92+00:00","citations":[{"url":"https://github.com/kwaipilot/SWE-Compass","path":null,"label":"kwaipilot/SWE-Compass","type":"external"},{"url":"https://github.com/kwaipilot/KAT-Coder-Agent","path":null,"label":"kwaipilot/KAT-Coder-Agent","type":"external"},{"url":"https://github.com/kwaipilot/KAT-Coder","path":null,"label":"kwaipilot/KAT-Coder","type":"external"},{"url":"https://github.com/kwaipilot/experiments","path":null,"label":"kwaipilot/experiments","type":"external"},{"url":"https://github.com/kwaipilot/SWE-Compass","path":null,"label":"kwaipilot/SWE-Compass","type":"external"},{"url":"https://github.com/kwaipilot/KAT-Coder-Agent","path":null,"label":"kwaipilot/KAT-Coder-Agent","type":"external"},{"url":"https://github.com/kwaipilot/KAT-Coder","path":null,"label":"kwaipilot/KAT-Coder","type":"external"},{"url":"https://github.com/kwaipilot/experiments","path":null,"label":"kwaipilot/experiments","type":"external"},{"url":"https://ai-tldr.dev/releases/kwai-keye-vl-2-0-30b-a3b/","path":null,"label":"ai-tldr.dev/releases","type":"external"},{"url":"https://www.ai-all.info/en/ai-models/keye-vl-2-0-30b-a3b","path":null,"label":"ai-all.info/en","type":"external"},{"url":"https://blog.csdn.net/kuaishoutech/article/details/161431079","path":null,"label":"blog.csdn.net/kuaishoutech","type":"external"},{"url":"https://www.aipuzi.cn/ai-news/keye-vl-2-0-30b-a3b.html","path":null,"label":"aipuzi.cn/ai-news","type":"external"},{"url":"https://36kr.com/newsflashes/3521399817051270","path":null,"label":"36kr.com/newsflashes","type":"external"},{"url":"https://www.36kr.com/p/1866758221189888","path":null,"label":"36kr.com/p","type":"external"}],"provenance":{"provider":"deepseek","model":"deepseek-v4-pro","workflow":"onlylabs-deepagents-analysis-v3","agent":"deepagents"},"evidence":{"total":14,"pages":4,"events":4,"web":6,"signal_desks":{"forks":1,"repos":3,"hiring":0,"talking":0,"releases":0},"data_radar_lanes":null,"data_radar_matches":null}},"signal_counts":{"total":4,"model_released":0,"release":0,"repo_new":3,"repo_forked":1,"post_published":0,"job_opened":0}}