Who's building their AI team — as of August 6, 2026
Companies ranked by number of active AI positions. Trend compares to previous weekly snapshot.
| # | Company | Open Roles | Change | Trend |
|---|---|---|---|---|
| 1 | Databricks | 137 | +3 | Growing |
| 2 | Stripe | 112 | -1 | Stable |
| 3 | Anthropic | 105 | +10 | Growing |
| 4 | Shield AI | 98 | +4 | Growing |
| 5 | Brex | 72 | -6 | Shrinking |
| 6 | Palantir | 67 | +5 | Growing |
| 7 | xAI | 67 | +1 | Stable |
| 8 | OpenAI | 58 | -3 | Shrinking |
| 9 | Scale AI | 54 | +4 | Growing |
| 10 | CoreWeave | 54 | -7 | Shrinking |
| 11 | Figma | 46 | +2 | Stable |
| 12 | Vercel | 17 | -3 | Shrinking |
| 13 | Sierra AI | 14 | +14 | New |
| 14 | Cohere | 12 | — | Stable |
| 15 | Airtable | 12 | +12 | New |
| 16 | Mercury | 10 | -1 | Stable |
| 17 | Harvey | 8 | +8 | New |
| 18 | SambaNova | 6 | +6 | New |
| 19 | Cognition | 6 | +6 | New |
| 20 | Writer | 5 | -7 | Shrinking |
| 21 | Modal | 5 | +5 | New |
| 22 | Crusoe Energy | 5 | +5 | New |
| 23 | Webflow | 5 | +5 | New |
Tools driving AI engineering demand right now. Updated August 6, 2026.
| Tool | Category | Hiring Signal | Intel |
|---|---|---|---|
| Claude Mythos | Foundation Model | 🔴 Very High | Leaked by Anthropic Apr 1. Described as "step change in capabilities" and strongest model on cybersecurity tasks. New Capybara pricing tier above Opus 4.6. Driving immediate demand for Anthropic ecosystem specialists and AI security researchers. |
| Claude Code | Autonomous Agent | 🔴 Very High | Claude Sonnet 5 is the default model (Free + Pro plans since Jul 1, 80.4% Terminal-Bench 2.1); Opus 4.7/4.8 available for the max-reasoning tier. Per "The Making of Claude Code" (Jul 9), 95% of Claude Code's own codebase is now written by Claude Code itself — forcing companies to hire Recursive AI Dev Auditors to review AI-written production code at scale. |
| Cursor | AI-Native IDE | 🔴 Very High | $2B+ ARR, ~$50B valuation (Apr 2026 fundraising). 360K+ paid users. Cursor Automations (event-driven agent pipelines) launched — new AI Workflow Automation Engineer role emerging. |
| Qwen3.5 Omni | Open-Source Multimodal | 🟠 High | Released Apr 1 by Alibaba. Native text + audio + video + realtime in a single pipeline — most capable open-weight multimodal to date. Competitive with GPT-4o. Driving demand for Multimodal AI Engineers. |
| GitHub Copilot | IDE Extension | 🟠 High | 26M users, agent mode GA across VS Code and JetBrains. 41% of production commits now include AI-generated lines (Sourcegraph, Mar 2026). Driving demand for AI QA Specialists. |
| Langflow | AI Pipeline Builder | 🟠 High CVE | CVE-2026-33017 on CISA KEV — patch required. Security engineer hiring up sharply at companies running Langflow in production. |
| Claude Sonnet 5 | Foundation Model | 🔴 Very High | Default model on Free + Pro plans since Jul 1, 2026 (80.4% Terminal-Bench 2.1, $2/M input promo through Aug 31). Driving demand for Sonnet 5 Migration Engineers auditing CLAUDE.md/model-routing for teams still hardcoded to older Claude versions. |
| Claude Fable 5 | Foundation Model | 🟠 High | Restored to Claude.ai, Claude Platform, Claude Code, and Claude Cowork on Jul 1, 2026 after a 19-day US export-control shutdown (Jun 12-30). Mythos 5 remains partially restricted to ~100 vetted orgs. The outage is driving demand for Sonnet 5 Migration Engineers who build fallback-model wiring so teams aren't hardcoded to a single frontier model when access can be pulled by regulators with days of notice. |
| GPT-5.6 (Sol / Terra / Luna) | Foundation Model | 🟠 High | OpenAI's latest frontier lineup, released Jul 9, 2026 in three variants (Sol, Terra, Luna). Enterprises benchmarking GPT-5.6 against Claude Sonnet 5 / Opus 4.8 for cost and quality tradeoffs are driving demand for Multi-Provider LLM Architects and Frontier Model Evaluators. |
| Grok 4.5 | Frontier Model | 🟠 High | SpaceX AI / xAI released Grok 4.5 on Jul 8, 2026 — Elon Musk labeled it "Opus-class," and it ranks #1 on the Harvey Legal Agent Benchmark, the third Opus-class model available (alongside Claude Opus 4.8 and Fable 5). Teams evaluating it as a Claude Code alternative are hiring Grok 4.5 Integration Engineers. |
| Devin Desktop (formerly Windsurf) | Autonomous Agent / IDE | 🟠 High | Cognition rebranded Windsurf to Devin Desktop on Jun 2, 2026, retiring the Cascade agent in favor of Devin Local (rewritten in Rust, ~30% more token-efficient, supports parallel subagents) and adding the open Agent Client Protocol (ACP) so Codex, Claude Agent, and Gemini CLI run as first-class agents in one editor. Cascade hit EOL Jul 1, 2026 — teams still wired to it are hiring Agent Orchestration Engineers to migrate. |
| Mistral Vibe | Unified Work + Code Agent | 🟠 High | Mistral folded Le Chat into Vibe on May 28, 2026 — one agent and one subscription spanning Work Mode (inbox, research, docs) and Code Mode (features, bug fixes, reviewable PRs), powered by Mistral Medium 3.5 (77.6% SWE-bench Verified) with remote agent sessions and VS Code/CLI integration. Free/Pro/Team/Enterprise tiers are driving demand for engineers who can wire Work Mode and Code Mode into existing enterprise tool stacks. |
| Z.ai ZCode (GLM-5.2) | Open-Weight Coding Agent | 🟠 High | Z.ai released GLM-5.2 on Jun 16, 2026 under an MIT license — a 744B-parameter MoE model (40B active) with a 1M-token context window, scoring 62.1 on SWE-bench Pro (ahead of GPT-5.5, behind Claude Opus 4.8). ZCode, Z.ai's native agentic IDE tuned for GLM-5.2, undercuts Claude Code and Cursor on price (GLM Coding Plan from ~$12.60/mo promo). Cost-sensitive teams evaluating open-weight alternatives are hiring AI Cost Optimization Engineers. |
| n8n | Automation Platform | 🟡 Medium CVE | CVE-2026-21858 CVSS 10.0 — patch required. MCP + n8n workflows driving AI Workflow Automation demand. |
| Microsoft Conductor | Multi-Agent Orchestrator | 🟠 High | May 2026 GA. Zero-token routing — Conductor dispatches sub-agents without consuming tokens for the orchestration layer itself. Direct LangChain/LangGraph competitor. Driving demand for Agent Orchestration Engineers and multi-agent architects who can benchmark Conductor vs LangChain for enterprise deployments. |
| Apple iOS 27 AI Platform | Mobile AI Platform | 🟠 High | Apple opening iOS 27 AI provider slots to Anthropic and Google (alongside OpenAI) — first multi-provider on-device AI architecture on iOS. Launch window May–Jun 2026. Driving demand for Mobile AI Engineers with iOS + LLM integration experience, and Anthropic/Claude API specialists who can optimize for on-device constraints. |
| xAI Grok Build | Autonomous Agent | 🟠 High | xAI open-sourced its Grok Build coding agent under Apache 2.0 in mid-Jul 2026, a direct response to a security incident in which the tool uploaded user repositories to a public Google Cloud bucket rather than a planned strategic move. A cautionary case study for teams handing agentic coding tools repo access — driving demand for AI Security Engineers who audit agent tool-permission scopes before adoption. |
| ChatGPT Work | AI Work Assistant | 🟠 High | OpenAI launched ChatGPT Work (powered by GPT-5.6) in mid-Jul 2026, combining chat with Codex-style coding so non-technical users can generate documents, presentations, and websites — positioned as cheaper and more broadly available than Anthropic's Claude Cowork, which shipped on iOS/Android the same day, removing its desktop-only constraint. Enterprises evaluating both are hiring Agent Orchestration Engineers to manage rollout and governance across coworker-style agents. |
Security vulnerabilities (CVE) in AI tools are increasing demand for AI Security Engineers. View AI Security jobs →
Know which companies are scaling AI teams before anyone else. Weekly data-driven intelligence.
Hiring velocity is computed from live job listings tracked by LLMHire across 170+ AI companies. Jobs are sourced from public ATS APIs (Greenhouse, Ashby, Lever) and refreshed 6× daily. "Change" compares current active listings to the previous weekly snapshot. Companies with fewer than 5 active listings are excluded. Trends improve over time as more weekly snapshots accumulate.