arize-phoenix: v12.35.0
12.35.0 (2026-02-09) Features add Claude Opus 4.6 model to playground (3554b76) Bug Fixes add timezone validation to log_spans_dataframe (45a07bb) anthropic streaming handling for new stop events (7ac1081) arize 8.0.0 inference schema (c42ce16) cost: update built-in model token prices (cd61bd9) deps: update arize-phoenix-client to 1.28.1 (b70f1c5) Improve...
Also shipped Feb 9, 2026Arize AI in February 2026 →
Git Sync now available in Public Preview
Git Sync in Grafana lets you manage your dashboards as code as JSON files stored in GitHub. You and your team can use it to version control, collaborate, and automate deployments efficiently. This feature, now starting to roll out to all Grafana Cloud instances in Public Preview , lets you benefit from a native Git workflow directly from Grafana.
Agents, Workers - Agents SDK v0.4.0: Readonly connections, MCP security improvements, x402 v2 migration, and custom MCP OAuth providers
The latest release of the Agents SDK ↗ brings readonly connections, MCP protocol and security improvements, x402 payment protocol v2 migration, and the ability to customize OAuth for MCP server connections. Readonly connections Agents can now restrict WebSocket clients to read-only access, preventing them from modifying agent state.
Agents - Interactive browser terminals in Sandboxes
The Sandbox SDK ↗ now supports PTY (pseudo-terminal) passthrough, enabling browser-based terminal UIs to connect to sandbox shells via WebSocket. sandbox.terminal(request) The new terminal() method proxies a WebSocket upgrade to the container's PTY endpoint, with output buffering for replay on reconnect.
AI Crawl Control - Analytics enhancements
AI Crawl Control metrics have been enhanced with new views, improved filtering, and better data visualization. Path pattern grouping In the Metrics tab > Most popular paths table, use the new Patterns tab that groups requests by URI pattern ( /blog/* , /api/v1/* , /docs/* ) to identify which site areas crawlers target most. Refer to the screenshot above.
AI Search now with more granular controls over indexing
Get your content updates into AI Search faster and avoid a full rescan when you do not need it. Reindex individual files without a full sync Updated a file or need to retry one that errored? When you know exactly which file changed, you can now reindex it directly instead of rescanning your entire data source.
Log Explorer - Tabs and pivots
Log Explorer now supports multiple concurrent queries with the new Tabs feature. Work with multiple queries simultaneously and pivot between datasets to investigate malicious activity more effectively. Key capabilities Multiple tabs: Open and switch between multiple query tabs to compare results across different datasets.
More from phoenixEvery phoenix release →
arize-phoenix: v12.34.0
12.34.0 (2026-02-06) Features evals: add missing tool_selection evaluator to both libraries (d6d1953) Bug Fixes add enum for new prompt span kind (b93abc5) cost: update built-in model token prices (860f703) normalize message_content.text to string before rendering in MessageContentListItem (d0908aa) Documentation evals: update built-in metric pages...
arize-phoenix: v13.0.0
13.0.0 (2026-02-13) ⚠ BREAKING CHANGES dataset evaluators Features Add autocomplete to LLM eval prompt editor (2dce02c) add available tools to experiment output (f17e359) add built-in LLM evaluator configs to graphql (4c8a70f) add custom providers to model menu (41fe933) Add dataset deep link after selection from playground (06c9687) add default output...
arize-phoenix-evals: v2.10.0
2.10.0 (2026-02-13) Features add tool response handling evaluator template (0738cb9) dataset evaluators (1274b03) dataset evaluators (5f47820) dataset evaluators (cf0e09a) enable full mustache prompt templates in the server (68ba0ed) filter built-in classification evaluators by label (aa95a3f) Bug Fixes templates: simplify template formatters (1f6df01)...
arize-phoenix-client: v1.29.0
1.29.0 (2026-02-13) Features dataset evaluators (5f47820) dataset evaluators (cf0e09a)
arize-phoenix-evals: v2.9.0
2.9.0 (2026-02-02) Features add FaithfulnessEvaluator and deprecate HallucinationEvaluator (fc8b1b5) add tool invocation accuracy metric (d1c89bf) evals: add missing tool_selection evaluator to both libraries (d6d1953) evals: add neutral optimization direction option (5768a14) return trace_id in Scores and improve trace decorator (a8dfb54) Bug Fixes Use...
arize-phoenix: v13.1.0
13.1.0 (2026-02-18) Features add model to evaluator details (cf48501) cli-starter-kit: use clack (e732d3e) Enhance table functionality with column resizing support in EvaluatorsTable and DatasetEvaluatorsTable (9264198) Implement column resizing in EvaluatorsTable and DatasetEvaluatorsTable (e7749ff) merge phoenix-sdk-python skill into phoenix-evals...
arize-phoenix: v13.2.0
13.2.0 (2026-02-19) Features add AWS Bedrock cross-region inference model prefix preference (affea10) Bug Fixes add parameter for deleting project after deleting experiment (57b7576) add sonnet 4.6 for aws bedrock model names (53ed96c) Revert lodash-es as it breaks production builds (e2574ca) Documentation add GTM tag to Mintlify config (83c9dda) document...
arize-phoenix: v13.3.0
13.3.0 (2026-02-20) Features evals: add conciseness classification evaluator (d43c4ee) Remove model inferences, dimensions, embeddings, and pointcloud UI (3dfa6ad) Bug Fixes Improve JSONDistanceEvaluator error messages (4e60cf5)