LLM observability meets application-level visibility.

Headline statistic. 40+ auto-instrumentations (source: github.com/Arize-ai/openinference, verified 2026-10-10) covering OpenAI, Anthropic, LangChain, LlamaIndex, and more. Phoenix delivers deep LLM-specific visibility.

Introduction. Arize Phoenix is an excellent LLM observability framework. 40+ auto-instrumentations (source: github.com/Arize-ai/openinference, verified 2026-10-10), best-in-class evaluation tools, and a free, self-hostable codebase (source-available; core under Elastic License 2.0). We respect that depth. Arize is now part of Dynatrace: on 2026-10-01 Dynatrace announced that it "has completed its acquisition of Arize" and that "Arize will continue supporting both Phoenix, its open-source platform, and AX, its enterprise platform" (source: Dynatrace press release, ir.dynatrace.com, dated 2026-10-01, verified 2026-10-10). This page is an honest look at where DeepCube takes a different approach, where Phoenix excels, and how the two can work together.

DeepCube is the product; spatial observability is the category. We are not an APM company, we are a spatial observability company.

Best-in-Class LLM Evaluation, Zero Application Context. When your LLM latency spikes, where is the bottleneck?

• 40+ auto-instrumentations (source: github.com/Arize-ai/openinference, verified 2026-10-10) covering OpenAI, Anthropic, LangChain, LlamaIndex, and more. Phoenix delivers deep LLM-specific visibility.
• But LLM calls do not execute in a vacuum. They depend on API gateways, vector databases, embedding services, and the application underneath.
• When your RAG pipeline slows down, is it the model, the retrieval layer, the database feeding context, or a noisy neighbor? Phoenix cannot answer that because it only sees the LLM layer.

See Your LLM Calls in Full System Context. Topology as your primary investigation surface.

• 3D spatial topology: Your LLM services, the APIs that call them, the vector databases they query, and the relationships between all of them.
• When something breaks, you see it in context: which service is affected, what it depends on, where the problem propagates.
• No context switching between your LLM tracing tool and your application monitoring. One environment. Application-level visibility.
• Phantoms: when a call goes out and nothing ever reports back from the other side, or a consumer goes quiet while messages keep arriving, DeepCube draws the missing side where it should be, so you see what is missing, not only what is there.

Architecture: How We Differ. Application-level correlation vs LLM-only tracing.

DeepCube provides spatial observability across distributed tracing, metrics, logs, and LLM observability in a single spatial environment. When your LLM latency spikes, DeepCube shows you whether the bottleneck is the model provider, your API gateway, the vector database, or an upstream dependency.

Phoenix shows you the LLM call. DeepCube shows you the LLM call in the context of the entire request lifecycle and the application dependencies around it.

Architecture table, full cells with sources.

• Scope, Arize Phoenix: LLM observability only (source, verified 2026-08-18)
• LLM Auto-Instrumentation, Arize Phoenix: 40+ auto-instrumentations (OpenInference) (source: github.com/Arize-ai/openinference, verified 2026-10-10)
• Service Topology, Arize Phoenix: No service topology (source, verified 2026-08-18)
• Application Metrics, Arize Phoenix: No application or infrastructure metrics. AX does ship custom metrics, dashboards and monitors on all tiers, but a custom metric is an AQL query over span and eval attributes, so latency, tokens, cost and eval labels, never service or host health (source, verified 2026-08-18)
• Cross-Signal Correlation, Arize Phoenix: LLM traces only (source, verified 2026-08-18)

AI: Tessa vs Alyx, PXI and Signal. Both ship AI agents. The difference is what they can see. (The AI table's DeepCube column was headed "Tessa (DeepCube)".)

Arize ships real AI agents: Alyx in AX, PXI in Phoenix (beta), and Signal, which proposes fixes on every plan; on Enterprise, Signal opens fix PRs against an attached GitHub repo. Working in your code is table stakes now, and Tessa does it too: it works in your connected codebase workspace and proposes the fix, and you review and own it. The difference is what the AI can see. When your RAG pipeline degrades, Tessa correlates the latency spike with the vector database query time and the application metrics around it, in one spatial view, because DeepCube carries the application telemetry Arize does not.

Phoenix and AX give you excellent LLM-layer visibility and agents that work inside it. DeepCube puts that LLM layer inside the whole application, and Tessa works across both.

AI table, full cells with sources.

• AI Diagnosis, Arize Phoenix: In-product AI diagnosis exists on both products but is scoped to the LLM layer. Alyx is AX's built-in AI engineering agent, covering trace investigation, aggregation, custom metrics, eval authoring, dataset curation and prompt iteration. PXI (Phoenix Intelligence) is the equivalent agent built into open-source Phoenix; its docs label it beta (verified 2026-10-10). Signal runs automatic failure-mode detection on a schedule, metered at 10 issues/mo Free and 25 Pro. With no service topology and no application signals, none of it correlates beyond the trace. Phoenix also serves a Remote MCP Server at /mcp exposing projects, traces, sessions, datasets, experiments, prompts and annotations to Claude Code, Cursor and VS Code; the older @arizeai/phoenix-mcp npm package is in maintenance mode (source, PXI, verified 2026-08-18)
• Codebase Access, Arize Phoenix: No codebase access on Free or Pro, which "cannot attach a repo, open fix PRs, or choose a custom AI provider". Arize's own docs describe PXI (beta) as "a coding agent pointed at your observability data instead of a codebase". On Enterprise, you can "attach a GitHub repo so Signal can open pull requests for proposed fixes" (Signal, PXI, verified 2026-10-10)
• What the AI can reach in your source, DeepCube: Tessa reads your code where it already lives, on your machine. No vendor GitHub App, no copy of your repo in our cloud, and only what she reads for your question goes to the model: a list of file names by default, and file contents only when she opens them. She works with your developer's own git access, so she can do only what that developer can already do, under their identity and your existing branch protections.
• What the AI can reach in your source, Arize Phoenix: On Enterprise, Signal attaches a GitHub repo with a personal access token: you create a GitHub skill ("name + personal access token") and "enter the repository as owner/repo" (source, verified 2026-10-10)
• Code Fixes, DeepCube: Available on every plan. Tessa traces the issue to the source file in your connected workspace and proposes the change. You review the change and own the outcome.
• Code Fixes, Arize Phoenix: Signal's investigations include a proposed fix, sometimes a code change, on every plan within the 10 and 25 issue allowances; acting on it is what is gated. Repo attach with fix PRs, Custom Managed Agents and custom AI providers are Enterprise-only. Alyx suggests prompt variants rather than editing your code, on all plans including Free (source, pricing, verified 2026-10-10)
• Accountability Model, Arize Phoenix: Agent-assisted within the LLM layer: Alyx in AX, PXI in Phoenix (beta), Signal on a schedule. On Enterprise, Signal opens fix PRs against an attached GitHub repo. No cross-signal accountability trail because there are no application signals to correlate (source, verified 2026-10-10)

Pricing: Application Observability Value. One platform vs LLM tool + APM tool + ops cost.

• One platform, not three: DeepCube includes LLM observability, application monitoring, and AI diagnosis. Phoenix covers the LLM layer only.
• No tool sprawl: Avoid managing separate tools for LLM tracing and APM. One bill. One UI. Full correlation.
• Predictable pricing: Nodes x tier price = monthly cost. No per-trace or per-evaluation charges.
• AI included: Tessa is part of every DeepCube plan at no extra charge. AI is included in the per-node price: there is no AI add-on required and no per-query AI charge. On paid plans the bundled Tessa allowance is a per-node budget; when it is spent, Tessa keeps working on a lighter model rather than stopping. Every plan can run every Tessa skill, including Code Fix and root-cause analysis; higher plans put a more capable model behind them.

Pricing table, full cells with sources.

• LLM Observability, Arize Phoenix: Free and self-hostable (source-available; core under Elastic License 2.0) / AX Pro from $50/mo (source: arize.com/pricing, verified 2026-10-10)
• Application Monitoring (APM), Arize Phoenix: Distributed tracing is AI-scoped only. OpenInference spans over LLM, tool, retriever and agent calls, with no service, host or infrastructure monitoring around them (requires a separate tool)
• AI Assistant, Arize Phoenix: Alyx optimizes prompts and investigates traces in-platform on all plans including Free, with no codebase access. Signal proposes fixes on every plan; repo attach, fix PRs and Custom Managed Agents are Enterprise-only (source, verified 2026-10-10)

DeepCube pricing from deepcube.ai/pricing. Arize Phoenix is source-available under the Elastic License 2.0; Arize platform pricing varies by plan. Verify current pricing at arize.com/pricing. All prices USD.

Where Phoenix Excels. Honest comparison: if LLM quality loops are your primary job, Phoenix earns its place.

• Best-in-class evaluation framework: hallucination, toxicity, relevance, and custom evals are Phoenix's core strength. A team whose main job is iterating on LLM output quality gets real depth here.
• 40+ OpenInference auto-instrumentations (source: github.com/Arize-ai/openinference, verified 2026-10-10) make it fast to capture rich LLM spans across OpenAI, Anthropic, LangChain, LlamaIndex, and more.
• Alyx (Arize AX), a built-in AI engineering agent (source, verified 2026-08-18) that investigates traces, authors evals and custom metrics, curates datasets and runs prompt experiments without leaving the platform, on every plan including Free. Phoenix has its own equivalent, PXI, in beta.

If your primary need is tightening the LLM prompt quality loop, you may legitimately prefer Phoenix's depth there. That is not a contradiction with spatial observability: DeepCube and Phoenix can be complementary. Keep Phoenix for evals, add DeepCube for full application context.

Already Using Phoenix? Add Application Context. OpenTelemetry makes it simple.

• Keep Phoenix for evals: If you love Phoenix's evaluation framework, keep using it. DeepCube adds the application monitoring layer Phoenix does not cover.
• Zero re-instrumentation: Phoenix is built on OpenTelemetry via OpenInference. Your existing OTel instrumentation works with both.
• Run both side by side: Compare the experience. When you see the value of correlating LLM performance with application health, you will understand why application-level context matters.
• Exit guarantee: If DeepCube is not right for you, change one endpoint URL. Your instrumentation stays exactly the same.