# agenta.ai > AI-optimized mirror of agenta.ai containing 50 pages totalling 26,687 words of clean markdown content, structured data, and semantic HTML. Original source: https://agenta.ai. Last updated: 2026-09-12T05:43:06.538Z. Each page is available as HTML (with JSON-LD structured data) and Markdown (text-only, ideal for LLMs and RAG). ## Homepage - [Agenta — The open-source workspace for your agents](/content/site-root.html): Agenta is the open-source workspace for your agents. Build agents through chat, improve them with feedback, and share them with your whole team — self-hosted or in the cloud. (375 words) ## Articles & Blog Posts - [Archive - Docs - Agenta](/content/docs/changelog/archive/index.html): Archive (787 words) - [Changelog - Docs - Agenta](/content/docs/changelog/index.html): New features, improvements, and fixes in Agenta., Stream, Stream, Stream, Stream, Stream, Stream (790 words) - [Top LLM Gateways 2025 — Agenta Blog](/content/blog/top-llm-gateways/index.html): We compare and test the top LLM gateways in 2025. These includes Litellm, Helicone, BricksLLM and Kong AI Gateway. (2,288 words) - [The AI Engineer's Guide to LLM Observability with OpenTelemetry — Agenta Blog](/content/blog/the-ai-engineer-s-guide-to-llm-observability-with-opentelemetry.html): Learn why LLM observability is critical for production AI. This guide covers traces, OpenTelemetry (OTel), and the LLMOps workflows you need to build reliable apps (3,474 words) - [8 Best Grok Bot Alternatives in 2026 — Agenta Blog](/content/blog/best-grok-bot-alternatives/index.html): Compare Grok Bot alternatives for one-off tasks, persistent AI coworkers, team work, personal assistance, model choice, and self-hosting. (2,650 words) - [Blog — Agenta](/content/blog/index.html): The latest updates and insights from Agenta — prompt management, evaluation, and observability for LLM apps. (1,274 words) - [Prompt Management for Non-Engineers: How Product Teams Can Own Their AI Prompts — Agenta Blog](/content/blog/prompt-management-for-non-engineers/index.html): How product managers and domain experts can contribute to AI prompt quality without writing code. A practical guide to collaborative prompt management. (2,000 words) - [Custom Code Evaluator - Docs - Agenta](/content/docs/1-0/evaluation/configure-evaluators/custom-evaluator/index.html): Write custom evaluators in Python, JavaScript, or TypeScript with access to inputs, outputs, and full trace data. (760 words) - [The Definitive Guide to Prompt Management Systems — Agenta Blog](/content/blog/the-definitive-guide-to-prompt-management-systems/index.html): Explore why prompt management is crucial for scaling AI applications from pilots to production. (1,146 words) - [Pricing — Agenta](/content/pricing/index.html): Simple pricing that scales with your team. Self-host the open-source platform for free, or run on Agenta Cloud with prompt management, evaluation, and observability built in. (974 words) - [Humanloop Sunsetting - Migration and Alternative — Agenta Blog](/content/blog/humanloop-sunsetting-migration-and-alternative/index.html): Humanloop has been acquired and goes offline on September 8, 2025. Agenta is an ideal alternative that lets you version prompts, evaluate, and monitor LLM apps easily. Migrate your prompts and workflows to Agenta with free white-glove migration support. (1,011 words) - [Prompt Playground - Docs - Agenta](/content/docs/1-0/prompt-engineering/playground/using-playground/index.html): Learn how to use Agenta's LLM playground for prompt engineering, model comparison, and deployment. A powerful alternative to OpenAI playground that supports multiple models and frameworks., Stream (1,054 words) - [Classification and Entity Extraction Evaluators - Docs - Agenta](/content/docs/1-0/evaluation/configure-evaluators/classification-entity-extraction/index.html): Agenta offers several evaluators to assess model performance in classification and entity extraction tasks. (532 words) - [Imprint — Agenta](/content/imprint/index.html): Legal imprint for Agentatech UG (haftungsbeschränkt), Berlin. Required by § 5 DDG. (67 words) - [Agenta on Mobile - Docs - Agenta](/content/docs/changelog/agenta-on-mobile/index.html): Use Agenta from your phone to chat with agents, handle tool approvals, return to past sessions, and manage your workspace., Stream (157 words, Aug 22, 2026) - [Agent Workspace - Docs - Agenta](/content/docs/changelog/agent-workspace/index.html): Organize and manage sessions more easily., Stream (329 words, Aug 12, 2026) - [Automation Runs - Docs - Agenta](/content/docs/changelog/automation-runs/index.html): Review automated work, provide human input, and improve future runs., Stream (310 words, Aug 12, 2026) - [API Keys Are Hidden from Agent Sandboxes - Docs - Agenta](/content/docs/changelog/api-keys-hidden-from-agent-sandboxes/index.html): Agents running in a cloud sandbox no longer see your provider API keys. Each key becomes a Daytona Secret scoped to a single host, and the sandbox holds only a placeholder. (595 words, Aug 4, 2026) - [Run Your Agents on Codex - Docs - Agenta](/content/docs/changelog/codex-harness/index.html): Codex is now a harness you can pick for an agent, alongside Claude Code and Pi. Five OpenAI models, local or cloud sandbox, tool approvals, and your own ChatGPT subscription when you self-host. (531 words, Aug 3, 2026) - [Shared Workspace Files - Docs - Agenta](/content/docs/changelog/file-attachments-in-agent-chat/index.html): One cloud folder for you, your team, and your agents. Upload images, PDFs, and whole folders, and the agent works from the same files you do. (526 words, Aug 2, 2026) - [Agenta Is Now a Workspace for Building Agents - Docs - Agenta](/content/docs/changelog/agenta-is-now-a-workspace-for-building-agents/index.html): Agenta is now a workspace for building and running agents: assistants made of instructions, tools, skills, permissions, and files that you can chat with or run in the background. (546 words, Jul 21, 2026) - [Dark Mode - Docs - Agenta](/content/docs/changelog/dark-mode/index.html): Agenta now ships a full dark theme. Switch between light, dark, and system from the top bar. (76 words, Jun 5, 2026) - [Annotation Queues - Docs - Agenta](/content/docs/changelog/annotation-queues/index.html): Run error analysis, collect SME feedback, bootstrap test sets from traces, and label test sets with rubrics or ground truth. (501 words, May 18, 2026) - [Webhooks and GitHub Automations for Prompt Deployments - Docs - Agenta](/content/docs/changelog/deployment-webhooks-and-github-automations/index.html): Trigger webhooks and GitHub Actions when you deploy a prompt. Use repository dispatch, workflow dispatch, or a custom HTTPS endpoint. (318 words, Mar 11, 2026) - [Enterprise Compliance Features - Docs - Agenta](/content/docs/changelog/enterprise-compliance-features/index.html): Multi-organization support, SSO with any OIDC provider, domain verification, and a new US region. (291 words, Feb 17, 2026) - [Chat Sessions in Observability - Docs - Agenta](/content/docs/changelog/chat-sessions-observability/index.html): Track and analyze multi-turn conversations with session grouping, cost analytics, and conversation flow visualization. (462 words, Jan 9, 2026) - [Jinja2 Template Support in the Playground - Docs - Agenta](/content/docs/changelog/jinja2-template-support/index.html): You can now use Jinja2 templates in your prompts. Jinja2 is available in both the Playground and in prompt management. (221 words, Nov 17, 2025) - [Agenta Core is Now Open Source - Docs - Agenta](/content/docs/changelog/open-sourcing-agenta/index.html): Agenta's core product is now open source under the MIT license. All functional features including evaluation, prompt management, and observability are available to the community. (259 words, Nov 13, 2025) - [Evaluation SDK - Docs - Agenta](/content/docs/changelog/evaluation-sdk/index.html): Run programmatic evaluations of complex AI agents and workflows from code. Evaluate agents built with any framework with full control over test data and evaluation logic. View results in the Agenta dashboard with traces and comparison views. (367 words, Nov 12, 2025) - [Customize LLM-as-a-Judge Output Schemas - Docs - Agenta](/content/docs/changelog/customize-llm-as-a-judge-output-schemas/index.html): Learn how to customize LLM-as-a-Judge evaluator output schemas with binary, multiclass, or custom JSON formats. Enable reasoning for better evaluation quality and structure feedback to match your workflow needs. (301 words, Nov 10, 2025) - [Documentation Architecture Overhaul - Docs - Agenta](/content/docs/changelog/documentation-architecture-overhaul/index.html): We've completely rewritten and restructured our documentation with a new architecture. This is one of the largest updates we've made, involving a near-complete rewrite of existing content. (142 words, Nov 3, 2025) - [Filtering Traces by Annotation - Docs - Agenta](/content/docs/changelog/filtering-traces-by-annotation/index.html): You can now filter and search traces based on their annotations. This helps you find traces with low scores or bad feedback quickly. (187 words, Oct 14, 2025) - [Deep URL Support for Sharable Links - Docs - Agenta](/content/docs/changelog/deep-url-support-for-sharable-links/index.html): URLs across Agenta now include workspace context, making them fully shareable between team members. Previously, URLs would always point to the default workspace, causing issues when refreshing pages or sharing links. (157 words, Sep 24, 2025) - [DSPy Integration - Docs - Agenta](/content/docs/changelog/dspy-integration/index.html): Trace and debug your DSPy applications with Agenta. (23 words, Aug 29, 2025) - [Annotate Your LLM Response (preview) - Docs - Agenta](/content/docs/changelog/annotate-your-llm-response-preview/index.html): One of the major feature requests we had was the ability to capture user feedback and annotations (e.g. scores) to LLM responses traced in Agenta. (121 words, May 15, 2025) - [Documentation Overhaul, New Models, and Platform Improvements - Docs - Agenta](/content/docs/changelog/documentation-overhaul-new-models-and-platform-improvements.html): We've made significant improvements across Agenta with a major documentation overhaul, new model support, self-hosting enhancements, and UI improvements. (178 words, May 2, 2025) - [Improvements to the Playground and Custom Workflows - Docs - Agenta](/content/docs/changelog/improvements-to-the-playground-and-custom-workflows.html): We've made several improvements to the playground, including: (59 words, Mar 19, 2025) - [Add Spans to Test Sets - Docs - Agenta](/content/docs/changelog/add-spans-to-test-sets/index.html): This release introduces the ability to add spans to test sets, making it easier to bootstrap your evaluation data from production. The new feature lets you:, Stream (79 words, Dec 11, 2024) - [Evaluator Testing Playground and a New Evaluation View - Docs - Agenta](/content/docs/changelog/evaluator-testing-playground-and-a-new-evaluation-view.html): Many users faced challenges configuring evaluators in the web UI. Some, Stream (136 words, Sep 22, 2024) - [Evaluators can access all columns - Docs - Agenta](/content/docs/changelog/evaluators-can-access-all-columns/index.html):