2025
Year in Review
We built with the community
9
hackathons hosted
We helped teams level up
44
in-person meetups for AI Engineers
We shared what we learned
29+
conferences worldwide
Then we shipped it
What the market made clear
Observability had to move beyond spans
→
LLM Observability for AI Agents
You can't judge agent actions in isolation
→
Session level evaluations
Prompt optimization should be continuous
→
Prompt learning
0
Total active teams on Arize AX
Arize helps AI teams operate with confidence at scale
3.5M+
Python downloads
monthly
1.8M+
container downloads
8k
GitHub stars
and
our OSS is loved by developers.
By The Numbers
Key Platform Metrics
567,811,213,028
📊
Ingested spans
387,112,008
⚡
Evaluations processed
42,925
🧪
Experiment runs
26,363
👥
Unique new users
84,507
🚀
New projects
A Year of Launches...
We have shipped a ton of new features this year.
91
🎉
Total Releases in 2025
January 2025
2 releases
Voice application tracing and evaluation
Customize colors on dashboards
February 2025
6 releases
OTEL tracing via HTTP
Create dataset from CSVs
Specify which columns of data to export when exporting traces
Schedule monitors
Expand and collapse data views
Labeling queues
March 2025
2 releases
Customize session view
Test online evaluation tasks in playground
April 2025
11 releases
Historical batch runs for evals
View task run history and errors
Prompt Hub Python SDK
New time selector including timezone settings
A/B test prompts in the playground
Improved function calling in the prompt playground
Cancel running tasks
Run evaluations on experiments
Run up to 100 rows in the playground
Security audit logs
Custom code evaluators
May 2025
6 releases
Latency and token counts in the playground
Column selection in the prompt playground
Attribute search on traces
Sleeker display of inputs and outputs on a span
Expanded OpenAI model support in prompt playground and tasks
Realtime trace ingestion for all Arize AX instances
June 2025
7 releases
Arize Alyx trace troubleshooting
Agent and multi-agent visualization
Session-level evaluations
Agent trajectory evaluations
Prompt learning
Playground views
Arize Database (ADB)
July 2025
8 releases
Cost tracking
Customizable hotkey for Alyx
Saved filters for traces
Arize tracing assistant
Prompt Hub release tags
Alyx build eval skill
Preview examples in evals UI
OpenInference Java
August 2025
16 releases
Experiment comparison & data visualization improvements
Image support for datasets and labeling queues
Revamped evals and tasks experience
Expanded annotation configuration capabilities
Trace interactivity improvements
Support for GPT-5 in prompt playground
Project UX enhancements
Alyx Copilot API advancements
Dedicated agent graph tab
Dataset filtering & REST API updates
Experiment refinements
Dataset management upgrades
Experiment trace improvements
Trace total cost views
Playground view defaults
Delete annotation queues
September 2025
12 releases
Dashboard improvements
Search in settings
Improved annotator selection for new labeling queues
Annotation config: optimization direction
Freely position dashboard widgets
Alyx synthetic data generation skill
Session and trace evaluations
Large dataset runs in playground
Dashboard rework enhancements
Improved sessions view
Autosave annotations
Experiment annotations on experiment compare page
October 2025
14 releases
API-driven monitors
Automatic threshold ranges for monitors
Data fabric
New timeline tab for traces
Tags
Sort datasets and experiment listing table
Support for tool call IDs
Add data region to login page
Add auth failures to tracing
Total traces on stats bar
Suppport for GPT OSS models on Bedrock
Expanded LLM support
Dashboard widget time setting
Autocomplete for annotations on dataets
November 2025
4 releases
Structured outputs support for playground
Session annotations
Integrations revamp
OpenInference TypeScript 2.0
December 2025
3 releases
Multi-span filters
Support for GPT-5.2 and GPT-5.2-Pro
Support for Opus & Haiku 4.5
Special thanks to
Our customers
and many more!
Our partners
and
our team! ❤️
Join us, we're hiring
🤙
