Vendor Status Tracker
This category consolidated hard. Of the 73 tools we cover, 8 are no longer viable choices and 10 have been acquired. Most comparison content still lists them as current.
| Tool | Status | Date | What happened |
|---|---|---|---|
Gentrace Eval Frameworks | Shut down | - | Company shut down; source released on GitHub under MIT. |
| LA Literal AI Observability & Tracing | Shut down | 31 Oct 2025 | Self-host image discontinued; cloud gone in 2026. Chainlit to community maintenance 1 May 2025. |
OpenAI Evals Eval Frameworks | Shut down | 30 Nov 2026 | Hosted platform read-only 31 Oct 2026, shuts down 30 Nov. OSS repo lightly maintained. |
| W WhyLabs Observability & Tracing | Shut down | - | Company discontinued operations and open-sourced the entire platform. |
LLM Guard Guardrails & Safety | Unmaintained | 9 Jul 2026 | Repository archived and read-only. Hugging Face models also unmaintained. |
Pezzo Prompt Management | Unmaintained | Jun 2025 | Last commit was a docs edit. 11 pull requests open. No archival notice. |
Martian LLM Gateways | Pivoted away | - | Company active but sells interpretability research. Router no longer presented as a product. |
Vellum Prompt Management | Pivoted away | - | Now sells a consumer personal AI assistant. No developer platform pricing published. |
Aporia Guardrails & Safety | Absorbed into platform | 23 Dec 2024 | Acquired by Coralogix, reported ~$50M. Founders lead Coralogix AI. |
| C CalypsoAI Guardrails & Safety | Absorbed into platform | 26 Sep 2025 | Acquired by F5. Announced at $180M, closed at $145.2M per F5's 10-Q. |
Prompt Security Guardrails & Safety | Absorbed into platform | 5 Aug 2025 | Acquired by SentinelOne. Press reported ~$250M; SEC filing records ~$180M. |
| B Baserun Prompt Management | Acquired | - | Acquired by LlamaIndex. Announced only as "(Acquired)" in its X bio. |
| | Acquired | Mar 2026 | Acquired by Mintlify, then placed in maintenance mode. No roadmap. |
Lakera Guardrails & Safety | Acquired | Sep 2025 | Acquired by Check Point, reported around $300M. Zurich team still maintains detection models. |
Promptfoo Eval Frameworks | Acquired | 9 Mar 2026 | Acquired by OpenAI, folding into OpenAI Frontier. Remains MIT. |
Traceloop Observability & Tracing | Acquired | Mar 2026 | Acquired by ServiceNow for AI Control Tower. OpenLLMetry stays Apache 2.0. |
TruLens Eval Frameworks | Acquired | May 2024 | TruEra acquired by Snowflake. Still MIT and self-hostable. |
W&B Weave Observability & Tracing | Acquired | - | Weights & Biases acquired by CoreWeave for a reported $1.7B. |
Agenta Prompt Management | Active | - | Independent and actively developed as of our last check. |
AgentOps Agent Evaluation | Active | - | Independent and actively developed as of our last check. |
Amazon Bedrock Evaluations Agent Evaluation | Active | - | Independent and actively developed as of our last check. |
Arize AX Observability & Tracing | Active | - | Independent and actively developed as of our last check. |
Arize Phoenix Observability & Tracing | Active | - | Independent and actively developed as of our last check. |
Arthur Guardrails & Safety | Active | - | Independent and actively developed as of our last check. |
Azure AI Foundry Evaluation Agent Evaluation | Active | - | Independent and actively developed as of our last check. |
Braintrust Eval Frameworks | Active | - | Independent and actively developed as of our last check. |
Cloudflare AI Gateway LLM Gateways | Active | - | Independent and actively developed as of our last check. |
Confident AI Eval Frameworks | Active | - | Independent and actively developed as of our last check. |
Confident AI (DeepEval) Eval Frameworks | Active | - | Independent and actively developed as of our last check. |
Databricks Agent Evaluation Agent Evaluation | Active | - | Independent and actively developed as of our last check. |
Datadog LLM Observability Observability & Tracing | Active | - | Independent and actively developed as of our last check. |
Evidently Eval Frameworks | Active | - | Independent and actively developed as of our last check. |
Fiddler AI Guardrails & Safety | Active | - | Independent and actively developed as of our last check. |
Freeplay Prompt Management | Active | - | Independent and actively developed as of our last check. |
Galileo Observability & Tracing | Active | - | Independent and actively developed as of our last check. |
Giskard Eval Frameworks | Active | - | Independent and actively developed as of our last check. |
Guardrails AI Guardrails & Safety | Active | - | Independent and actively developed as of our last check. |
HiddenLayer Guardrails & Safety | Active | - | Independent and actively developed as of our last check. |
HoneyHive Observability & Tracing | Active | - | Independent and actively developed as of our last check. |
Inspect AI Eval Frameworks | Active | - | Independent and actively developed as of our last check. |
Kong AI Gateway LLM Gateways | Active | - | Independent and actively developed as of our last check. |
Laminar Observability & Tracing | Active | - | Independent and actively developed as of our last check. |
Langfuse Observability & Tracing | Active | - | Independent and actively developed as of our last check. |
LangSmith Observability & Tracing | Active | - | Independent and actively developed as of our last check. |
Langtail Prompt Management | Active | - | Independent and actively developed as of our last check. |
Langtrace Observability & Tracing | Active | - | Independent and actively developed as of our last check. |
LangWatch Agent Evaluation | Active | - | Independent and actively developed as of our last check. |
Latitude Prompt Management | Active | - | Independent and actively developed as of our last check. |
LiteLLM LLM Gateways | Active | - | Independent and actively developed as of our last check. |
LM Evaluation Harness Eval Frameworks | Active | - | Independent and actively developed as of our last check. |
Lunary Observability & Tracing | Active | - | Independent and actively developed as of our last check. |
Maxim AI Agent Evaluation | Active | - | Independent and actively developed as of our last check. |
MLflow Observability & Tracing | Active | - | Independent and actively developed as of our last check. |
New Relic AI Monitoring Observability & Tracing | Active | - | Independent and actively developed as of our last check. |
Not Diamond LLM Gateways | Active | - | Independent and actively developed as of our last check. |
NVIDIA NeMo Guardrails Guardrails & Safety | Active | - | Independent and actively developed as of our last check. |
Openlayer Eval Frameworks | Active | - | Independent and actively developed as of our last check. |
OpenLIT Observability & Tracing | Active | - | Independent and actively developed as of our last check. |
OpenRouter LLM Gateways | Active | - | Independent and actively developed as of our last check. |
Opik Observability & Tracing | Active | - | Independent and actively developed as of our last check. |
Parea AI Prompt Management | Active | - | Independent and actively developed as of our last check. |
Patronus AI Eval Frameworks | Active | - | Independent and actively developed as of our last check. |
Portkey LLM Gateways | Active | - | Independent and actively developed as of our last check. |
PromptHub Prompt Management | Active | - | Independent and actively developed as of our last check. |
PromptLayer Prompt Management | Active | - | Independent and actively developed as of our last check. |
Pydantic Logfire Observability & Tracing | Active | - | Independent and actively developed as of our last check. |
Ragas Eval Frameworks | Active | - | Independent and actively developed as of our last check. |
Sentry Observability & Tracing | Active | - | Independent and actively developed as of our last check. |
SigNoz Observability & Tracing | Active | - | Independent and actively developed as of our last check. |
TrueFoundry LLM Gateways | Active | - | Independent and actively developed as of our last check. |
UpTrain Eval Frameworks | Active | - | Independent and actively developed as of our last check. |
Vercel AI Gateway LLM Gateways | Active | - | Independent and actively developed as of our last check. |
Vertex AI Gen AI Evaluation Service Agent Evaluation | Active | - | Independent and actively developed as of our last check. |
Why this page exists
Directory sites and comparison articles are almost never pruned. A vendor only disappears from them if somebody actively removes it, and nobody is paid to do that. Open-source projects are worse still, because there is no shutdown event at all - the repository still loads, the README still reads well, and nothing tells you development stopped eighteen months ago.
Every entry here was checked against the vendor's own channels. Where we could not confirm something, we say so on the tool's page rather than rounding to a verdict.
Frequently Asked Questions
How do you verify these?
Against the company's own channels wherever possible - a banner on their site, a repository archival notice, an SEC filing, a press release. Comparison articles and directory listings are not sufficient, because they are almost never pruned. Where we could not confirm something we say so rather than inferring it. Martian, for example, is listed as pivoted rather than shut down, because its homepage no longer sells a router but no discontinuation was ever announced.
Why does it matter if a tool was acquired rather than shut down?
Because the outcomes differ enormously and the label alone does not tell you which you are getting. Lakera was acquired by Check Point and its original team still maintains the detection models. Helicone was acquired by Mintlify and placed in maintenance mode with no roadmap. Protect AI was acquired by Palo Alto and its open-source LLM Guard was archived roughly a year later. Same word, three very different consequences for anyone who had built on them.
How do I check this myself for a tool you do not cover?
Look at the company's own homepage and ask what it currently sells, not what a directory says it sells. For open-source projects, check the date of the last substantive commit, whether pull requests are being merged or accumulating, and whether issues get responses. A polished README with eleven unmerged pull requests is a specific and common failure pattern. The absence of shutdown news is not evidence a vendor is still independent.
Why is a permissive licence not protection?
Because a licence protects your right to the code, not whether anyone keeps improving it. LLM Guard was MIT licensed, and MIT preserved the code perfectly while the project was archived and its detection models stopped being updated. For a guardrail, where the value lives in continuously updated detection rather than in the source, that distinction is the entire thing. An archived guardrail keeps returning confident passes while missing every attack technique invented after it stopped.