Opik 2.2.133 adds reasoning-effort selection to online evaluations
Updated Opik
Summary
Evaluation controls and annotation workflow changes are grouped with the preceding maintenance release.
- Tool
- Opik
- Category
- Agent infra
- Date
- Pricing
- Not stated in the sources we cited.
Opik 2.2.133 lets online-evaluation rules choose reasoning effort for OpenAI and OpenRouter. This makes a model setting part of evaluation configuration; it does not establish that increasing effort improves an application's scores.
The annotator card and preview are simplified, and the workflow can continue the same Ollie chat. The backend also adjusts address failover without replaying requests.
Version 2.2.132, released earlier in the same window, is included here rather than counted as a separate launch. It moves the relevant frontend, SDK and CI tooling to Node 26, adds a seven-day minimum npm release age, and includes dependency and trace-read changes.
For existing users, the concrete checks are whether online evaluation rules preserve the intended model settings and whether local development or CI uses the required runtime. Provider access and model charges are separate from configuring an evaluator.
See the opik hub.
More from Opik
Opik adds reusable playground variants and tightens provider logging
OpikFour related releases combine prompt-variant controls with logging, retry and cost-accounting corrections.
Category: Agent infra
Pricing: Not stated
Opik 2.2.122–127 strengthens execution and request controls
OpikA grouped release update covers sandboxed prompt rendering, destination policies and opt-in executor isolation.
Category: Agent infra
Pricing: Not stated
Opik updates human annotation and evaluation workflows
OpikVersions 2.2.120–121 add custom annotator views, thread environment filtering and playground table controls.
Category: Agent infra
Pricing: Not stated
Opik 2.2.119 moves playground experiments to the server
OpikRegular-dataset playground experiments can run and stop server-side, with fixes for results, retries and tool-call tracing.
Category: Agent infra
Pricing: Not stated
Opik 2.2.104–2.2.105 adds dataset history viewing and Cohere tracing
OpikOpik adds read-only access to older dataset and test-suite versions, a native Cohere SDK integration, and provider-aware usage pricing.
Category: Agent infra
Pricing: Not stated
Opik 2.2.103 adds native Together AI tracing integration
OpikThe October 16 update adds a Together AI SDK integration and fixes Google ADK span tracking and provider error handling.
Category: Agent infra
Pricing: Open source
Opik 2.2.97 fixes Anthropic max_tokens and tiered storage
OpikOpik 2.2.97, released October 12, 2026, honours max_tokens for Anthropic chat completions and adds tiered-storage metrics and Helm changes.
Category: Agent infra
Pricing: Open source
Questions
What is Opik?
According to its docs, Opik is an open-source platform that helps you understand what your LLM application is doing, measure how well it works, and improve it.
Is Opik open source?
Yes. The docs say Opik is open source, and the GitHub API reports the repository license as Apache-2.0.
Does Opik have an MCP server?
The docs say you can install the Opik MCP server to drive your workspace from tools such as Claude Code, Cursor, VS Code Copilot, Codex or opencode.
Filed on Agent Radar. Claims in this note are limited to the sources above.