Keep your OpenInference traces. Add security and compliance.
Arize Phoenix is a great OpenInference tracing + evaluation tool. EvalGuard imports your Phoenix spans — model, provider, token counts, and input/output all carry over — and puts a 300+-plugin red-team platform, a real-time input + output firewall, a BYOK gateway, and a SOC 2 evidence engine on the same data. No sign-up needed to run your first import.
Honest positioning
Where Phoenix stops, EvalGuard keeps going
Phoenix is a first-class tracing + eval tool built on the OpenInference standard, and self-hostable — though its server is Elastic-licensed rather than open source. EvalGuard overlaps on tracing and evals, then extends into the hosted security platform, runtime firewall, AI gateway, and compliance work you'd otherwise buy separately.
| Capability | Arize Phoenix | EvalGuard |
|---|---|---|
| OpenInference / OTel tracing | Yes | Yes — OTLP + OpenInference ingestion, governed |
| Evaluation + LLM-as-judge | Yes | Yes — 200+ scorers (LLM-as-judge, pairwise, rubric) |
| Span-level token / cost capture | Yes | Yes — per-org cost ledger + budgets |
| Licence & self-hosting | Self-hostable; server is Elastic License 2.0 (source-available), clients Apache-2.0 | Yes — Apache-2.0 core, self-host available |
| Red-team & security scans | Not offered | Yes — 300+ attack plugins |
| Runtime LLM firewall / guardrails | “Phoenix doesn’t block requests — your app does”; runtime guards live in Arize AX | Yes — real-time input + output firewall |
| AI gateway / BYOK proxy | “Phoenix is not an AI gateway” — but a playground chat-completions proxy with key custody exists | Yes — BYOK gateway, 15 proxied providers, similarity response cache |
| SOC 2 evidence automation | Not a documented product feature | Yes — live evidence engine + audit log |
Arize Phoenix claims on this page were verified against their source — github.com/Arize-ai/phoenix @ b4d9b19e6 (2026-08-08) — on 2026-08-10. The “Not offered” row was settled by searching that tree for an attack generator or a jailbreak/injection detector across the Python and TypeScript packages; the only attack list lives in an unpackaged tutorial notebook. Capabilities move; check their current docs before you decide. “SOC 2 evidence automation” describes a product capability, not any vendor’s own certification status.
This column describes Phoenix — not Arize AX, the commercial product, which does ship runtime guards.
Migration path
Bring your spans in one command
Everything stays local until you choose to run it — we never touch your Phoenix instance. Export your spans, then convert them to neutral-shape spans with the EvalGuard CLI.
# Python: from phoenix.client import Client
# df = Client().spans.get_spans_dataframe(project_identifier='default')
# df.to_json('phoenix-spans.json', orient='records', date_format='iso')npx @evalguard/cli import:traces --from phoenix phoenix-spans.json --output spans.jsonnpx @evalguard/cli init --template security-scan # scaffolds a scan config
npx @evalguard/cli scan:local <that-config> # no API key neededEvery OpenInference attribute — llm.model_name, token counts, and input/output.value — maps over (both the flattened dataframe and nested CLI-raw exports). Then layer on red team, a runtime firewall, and a SOC 2 evidence engine.
Want a hand with the migration?
Send us your Phoenix export and we'll help you map it and validate the first import. Free.