| 3.10.6 (opens in API7 Gateway docs) | 2026-08-25 | Feature | Added three metrics for Layer 4 traffic, joining the apisix_stream_connection_total counter that was already exported: apisix_stream_status counting completed sessions by termination status, listening address and upst… (opens in API7 Gateway docs) |
| 3.10.5 (opens in API7 Gateway docs) | 2026-08-11 | Fix | Fixed issue: Expired entries in the shared dictionary that backs the metrics were not reclaimed, so its usage only ever grew and it eventually stayed full, at which point new metric series were dropped. Expired entrie… (opens in API7 Gateway docs) |
| 3.10.5 (opens in API7 Gateway docs) | 2026-08-11 | Feature | The plugin set selected by global rules is now computed once per request and reused across phases, instead of being recomputed in each of them. On a deployment that enables the Prometheus plugin through a global rule,… (opens in API7 Gateway docs) |
| 3.10.4 (opens in API7 Gateway docs) | 2026-07-27 | Fix | Fixed issue: When the shared dictionary used for metrics filled up, the gateway could enter a loop that held a worker at 100% CPU and did not recover after traffic stopped. A full dictionary now degrades gracefully, l… (opens in API7 Gateway docs) |
| 3.10.1 (opens in API7 Gateway docs) | 2026-06-15 | Feature | Added LLM observability metrics: per-request prompt-token and completion-token distribution histograms (apisix_llm_prompt_tokens_dist, apisix_llm_completion_tokens_dist), a total-latency histogram, and time-to-first-t… (opens in API7 Gateway docs) |
| 3.10.1 (opens in API7 Gateway docs) | 2026-06-15 | Feature | Added mcp_request_type and mcp_tool_name labels to the http_status, http_latency, and bandwidth metrics, so MCP (Model Context Protocol) traffic can be broken down by request type (tools/list or tools/call) and by too… (opens in API7 Gateway docs) |
| 3.10.1 (opens in API7 Gateway docs) | 2026-06-15 | Feature | Reduced per-request overhead in the metrics logging phase by caching the disabled-labels map instead of rebuilding it on every request. (opens in API7 Gateway docs) |
| 3.9.19 (opens in API7 Gateway docs) | 2026-08-26 | Feature | Added three metrics for Layer 4 traffic, joining the apisix_stream_connection_total counter that was already exported: apisix_stream_status counting completed sessions by termination status, listening address and upst… (opens in API7 Gateway docs) |
| 3.9.18 (opens in API7 Gateway docs) | 2026-08-12 | Fix | Fixed issue: Expired entries in the shared dictionary that backs the metrics were not reclaimed, so its usage only ever grew and it eventually stayed full, at which point new metric series were dropped. Expired entrie… (opens in API7 Gateway docs) |
| 3.9.18 (opens in API7 Gateway docs) | 2026-08-12 | Feature | The plugin set selected by global rules is now computed once per request and reused across phases, instead of being recomputed in each of them. On a deployment that enables the Prometheus plugin through a global rule,… (opens in API7 Gateway docs) |
| 3.9.17 (opens in API7 Gateway docs) | 2026-07-28 | Fix | Fixed issue: When the shared dictionary used for metrics filled up, the gateway could enter a loop that held a worker at 100% CPU and did not recover after traffic stopped. A full dictionary now degrades gracefully, l… (opens in API7 Gateway docs) |
| 3.9.14 (opens in API7 Gateway docs) | 2026-06-15 | Feature | The Prometheus plugin now adds MCP tool dimensions (mcp_tool_name and mcp_request_type) to HTTP metrics, so MCP tools/call traffic can be broken down by tool and request type. (opens in API7 Gateway docs) |
| 3.9.14 (opens in API7 Gateway docs) | 2026-06-15 | Feature | The Prometheus plugin now exports LLM token-distribution histograms (apisix_llm_prompt_tokens_dist and apisix_llm_completion_tokens_dist) and an apisix_llm_latency histogram that includes time-to-first-token (type="tt… (opens in API7 Gateway docs) |
| 3.9.10 (opens in API7 Gateway docs) | 2026-04-22 | Feature | Prometheus, OpenTelemetry, Zipkin: Added a new response_source label (Prometheus) and apisix.response_source span attribute (OpenTelemetry, Zipkin) that classifies each response as "apisix" (generated by APISIX, such… (opens in API7 Gateway docs) |
| 3.9.8 (opens in API7 Gateway docs) | 2026-04-07 | Fix | Fixed issue: The apisix_llm_active_connections Prometheus gauge was never decremented when a plugin exited early via ngx.exit(), causing the metric to grow indefinitely and report incorrect active connection counts. (opens in API7 Gateway docs) |
| 3.9.7 (opens in API7 Gateway docs) | 2026-03-25 | Fix | Fixed issue: apisix_llm_* metrics were exported for all API routes, even those without AI plugins enabled, causing unnecessary metric cardinality and storage overhead. Also added disabled_labels support for LLM metric… (opens in API7 Gateway docs) |
| 3.8.21 (opens in API7 Gateway docs) | 2025-12-23 | Fix | Fixed issue: Optimized the performance after enabling the Prometheus plugin. (opens in API7 Gateway docs) |
| 3.5.0 (opens in API7 Gateway docs) | 2025-01-27 | Fix | Fixed issue: Even after the Prometheus plugin is disabled, all features related to Prometheus are not entirely shut down.. (opens in API7 Gateway docs) |
| 3.5.0 (opens in API7 Gateway docs) | 2025-01-27 | Feature | Supported custom configuration of DP metrics label through Prometheus plugin metadata. (opens in API7 Gateway docs) |
| 3.4.0 (opens in API7 Gateway docs) | 2025-01-07 | Feature | Supported custom configuration of DEFAULT_BUCKETS. (opens in API7 Gateway docs) |
| 3.2.16.5 (opens in API7 Gateway docs) | 2024-11-21 | Fix | Fixed issue: Enabling the prefer_name option in the Prometheus plugin will cause the filters on the monitor page to malfunction. (opens in API7 Gateway docs) |
Docs