{"slug": "ai-agents-weekly-deepseek-harness-deepseek-v4-pro-grok-bot-glm-5-3-gemini-3-7", "title": "🤖 AI Agents Weekly: DeepSeek Harness, DeepSeek-V4-Pro, Grok Bot, GLM-5.3, Gemini 3.7 Flash, Muse Glimmer, Harness Evolution Papers, and More", "summary": "DeepSeek open-sourced DeepSeek Harness v0.1 under MIT, a plugin-based agent harness built on the Cordis meta-framework, and launched DeepSeek V4-Pro-0813 with agentic benchmark scores including 87.9 on Terminal Bench 2.1 and 62.7 on DeepSWE, plus new API pricing effective August 16 with off-peak rates 50% below peak.", "body_md": "# 🤖 AI Agents Weekly: DeepSeek Harness, DeepSeek-V4-Pro, Grok Bot, GLM-5.3, Gemini 3.7 Flash, Muse Glimmer, Harness Evolution Papers, and More\n\n### DeepSeek Harness, DeepSeek-V4-Pro, Grok Bot, GLM-5.3, Gemini 3.7 Flash, Muse Glimmer, Harness Evolution Papers, and More\n\nIn today’s issue:\n\nDeepSeek open-sources its agent harness\n\nDeepSeek-V4-Pro ships agent upgrades\n\nxAI launches Grok Bot teammates\n\nZ.ai drops GLM-5.3 for coding\n\nGemini 3.7 Flash halves coding cost\n\nMeta open-sources Muse Glimmer\n\nGrok 4.6 hits frontier at half price\n\nZed launches Delta for agent teams\n\nEvo-Bench measures harness evolution\n\nStudy finds 91.8% of skills defective\n\nAnd all the top AI dev news, papers, and tools.\n\n## Top Stories\n\n### DeepSeek Open-Sources Its Agent Harness\n\nDeepSeek released DeepSeek Harness v0.1 as a developer preview, open-sourcing the codebase under MIT and opening it to anyone building agent harnesses.\n\n**Everything is a plugin:** The harness is built on the Cordis meta-framework, a kernel that mounts, unmounts, and resolves dependencies for models, tools, skills, sessions, sandboxes, storage, loops, scheduling, and UI as independent plugins.**Append-only session log:** Everything the model sees is recorded, so sessions can be resumed, forked, searched, and replayed rather than reconstructed from chat history.**Four runtime modes:** Standard ships the full toolset, Code orchestrates operations through TypeScript, Minimal strips down for benchmark runs, and Creator is for building custom presets.**Install path:** Runs via`npx @deepseek-ai/dsh web`\n\nor from source, and the repo has already cleared 93,000 stars.\n\n### DeepSeek Launches V4-Pro\n\nDeepSeek shipped V4-Pro-0813, a general availability release centered almost entirely on agent workloads.\n\n**Agentic benchmarks:** 87.9 on Terminal Bench 2.1, 62.7 on DeepSWE, 74.1 on Toolathlon-Verified, 83.3 on CyberGym, and 31.8 on public AutomationBench, tested through DeepSeek Harness in minimal mode.**Flexible reasoning effort:** Low, high, and max tiers across V4-Pro and V4-Flash let you dial spend per task instead of paying reasoning cost on trivial calls.**Native Responses API:** Ships OpenAI Responses API support with one-click Codex setup, and model names stay unchanged so existing integrations keep working.**Peak and off-peak pricing:** New API rates take effect August 16, with off-peak rates 50% below peak for schedulable batch and agent workloads.", "url": "https://wpnews.pro/news/ai-agents-weekly-deepseek-harness-deepseek-v4-pro-grok-bot-glm-5-3-gemini-3-7", "canonical_source": "https://nlp.elvissaravia.com/p/ai-agents-weekly-deepseek-harness", "published_at": "2026-08-15 16:43:19+00:00", "updated_at": "2026-08-15 17:10:56.666028+00:00", "lang": "en", "topics": ["ai-agents", "ai-infrastructure", "ai-products"], "entities": ["DeepSeek", "DeepSeek Harness", "DeepSeek V4-Pro-0813", "Cordis", "OpenAI Responses API"], "alternates": {"html": "https://wpnews.pro/news/ai-agents-weekly-deepseek-harness-deepseek-v4-pro-grok-bot-glm-5-3-gemini-3-7", "markdown": "https://wpnews.pro/news/ai-agents-weekly-deepseek-harness-deepseek-v4-pro-grok-bot-glm-5-3-gemini-3-7.md", "text": "https://wpnews.pro/news/ai-agents-weekly-deepseek-harness-deepseek-v4-pro-grok-bot-glm-5-3-gemini-3-7.txt", "jsonld": "https://wpnews.pro/news/ai-agents-weekly-deepseek-harness-deepseek-v4-pro-grok-bot-glm-5-3-gemini-3-7.jsonld"}}