Telluvian vs Galileo: hallucination detection compared
Telluvian and Galileo both detect hallucinations, by different methods. Telluvian reads the internal activations of a proxy model with probing classifiers and returns a score for every token, in the same API call that generates and routes the answer. Galileo, now Splunk Agent Observability after Cisco's acquisition, evaluates outputs with evaluator models (LLM judges or its Luna-2 small language models) as part of a wider observability platform.
Last reviewed
At a glance
| Telluvian | Galileo | |
|---|---|---|
| Model selection method | Prompt-aware. Model Select predicts each model’s performance on the specific prompt and picks the cheapest one that clears your xPerf quality bar. | Not a model router. It evaluates the outputs of models you call yourself. |
| Models and providers | 190+ models from OpenAI, Anthropic, Google, Qwen, DeepSeek, xAI, Z-AI, Moonshot and Meta, behind one OpenAI-compatible API. | Not applicable. An evaluation platform, not a gateway to models. |
| Pricing model | Prepaid, pay as you go. Provider list price for tokens, plus $0.05 per 1M input tokens for Model Select and $1.00 per 1M completion tokens for hallucination scores (optional). | From $100 a month per 1.2 million spans, billed annually, including 500 million Luna tokens a year. A free edition through the Splunk Observability Cloud free tier. |
| Hallucination detection | Built in. Per-token hallucination scores from probes reading an open-weight proxy model, including for closed models. Off by default. | Yes. Context Adherence (closed-domain hallucinations against the context you provide) and Correctness metrics, computed by LLM judges or Luna-2 small language models built for real-time use. |
Key differences
Method
Telluvian reads a proxy model's internal activations; Galileo asks evaluator models to judge the text.
Granularity
Telluvian scores each token, so you can see which words to check.
Reference context
Telluvian's probe needs none; Galileo's Context Adherence measures answers against context you supply.
Pricing
Telluvian charges per completion token with no subscription; Splunk Agent Observability starts at $100 a month per 1.2 million spans, billed annually.
Frequently asked questions
Is Galileo now part of Splunk?
Yes. Cisco has acquired Galileo, and the product is now called Splunk Agent Observability.
How do the detection methods differ?
Galileo computes metrics such as Context Adherence and Correctness with evaluator models: LLM judges or its own Luna-2 small language models. Telluvian uses probing classifiers that read the internal activations of an open-weight proxy model as the response is replayed through it, giving a score for each token, including for closed models.
Do I need retrieved context to detect hallucinations?
Not with Telluvian: the probe needs no reference documents and scores the model's own response from the proxy model's internal state. Galileo's Context Adherence does need it, because it measures whether the answer stays within the context you supplied.
Which is cheaper?
Splunk Agent Observability starts at $100 a month per 1.2 million spans, billed annually. Telluvian charges $1.00 per 1M completion tokens for hallucination scores, only on requests where you switch them on, with no subscription or commitment.
How does each handle legal work?
Legal is Telluvian's first market: a per-token score points at the exact span to check, such as a fabricated case citation or a misquoted statute. Galileo is a general observability platform covering many kinds of application.
Try it on your own prompts
Point an OpenAI SDK at Telluvian, send telluvian/gallery-1 as the model, and see which model answers each request. Questions about your use case go straight to the team.
Sources
Competitor details come from their own public docs and pricing pages, checked on . Where a page did not say, neither do we. Spotted something out of date? Tell us at hello@telluvian.ai.