harn_vm/llm/
model_test.rs

1use std::time::Instant;
2
3use serde::Serialize;
4
5use super::api::{
6    vm_call_llm_full_streaming, LlmCallOptions, LlmRoutePolicy, OutputFormat, ThinkingConfig,
7};
8use crate::value::{VmError, VmValue};
9
10const SMOKE_TEST_MAX_TOKENS: i64 = 32;
11
12#[derive(Clone, Debug, PartialEq, Eq)]
13pub struct ModelSmokeTestOptions {
14    pub model: String,
15    pub provider: Option<String>,
16    pub prompt: String,
17}
18
19#[derive(Clone, Debug, PartialEq, Serialize)]
20pub struct ModelSmokeTestResult {
21    pub model_id: String,
22    pub provider: String,
23    pub latency_ms: u64,
24    #[serde(skip_serializing_if = "Option::is_none")]
25    pub first_token_ms: Option<u64>,
26    pub input_tokens: i64,
27    pub output_tokens: i64,
28    pub estimated_cost_usd: f64,
29}
30
31pub async fn run_model_smoke_test(
32    options: ModelSmokeTestOptions,
33) -> Result<ModelSmokeTestResult, String> {
34    super::provider::register_default_providers();
35
36    let resolved = crate::llm_config::resolve_model_info(&options.model);
37    let model_id = resolved.id;
38    let provider = options
39        .provider
40        .as_deref()
41        .map(str::trim)
42        .filter(|provider| !provider.is_empty())
43        .map(str::to_string)
44        .unwrap_or(resolved.provider);
45    let api_key = super::helpers::resolve_api_key(&provider).map_err(vm_error_message)?;
46
47    if let Some(def) = crate::llm_config::provider_config(&provider) {
48        if super::supports_model_readiness_probe(&def) {
49            let readiness =
50                super::probe_openai_compatible_model(&provider, &model_id, &api_key).await;
51            if readiness.category == "model_missing" || readiness.category == "invalid_url" {
52                return Err(readiness.message);
53            }
54        }
55    }
56
57    let opts = LlmCallOptions {
58        provider: provider.clone(),
59        model: model_id.clone(),
60        api_key,
61        route_policy: LlmRoutePolicy::Manual,
62        fallback_chain: Vec::new(),
63        route_fallbacks: Vec::new(),
64        routing_decision: None,
65        session_id: None,
66        messages: vec![serde_json::json!({
67            "role": "user",
68            "content": options.prompt,
69        })],
70        system: None,
71        transcript_summary: None,
72        max_tokens: SMOKE_TEST_MAX_TOKENS,
73        temperature: None,
74        top_p: None,
75        top_k: None,
76        logprobs: false,
77        top_logprobs: None,
78        stop: None,
79        seed: None,
80        frequency_penalty: None,
81        presence_penalty: None,
82        output_format: OutputFormat::Text,
83        response_format: None,
84        json_schema: None,
85        output_schema: None,
86        output_validation: None,
87        thinking: ThinkingConfig::Disabled,
88        anthropic_beta_features: Vec::new(),
89        vision: false,
90        tools: None,
91        native_tools: None,
92        tool_choice: None,
93        tool_search: None,
94        cache: false,
95        timeout: None,
96        idle_timeout: None,
97        stream: true,
98        provider_overrides: None,
99        budget: None,
100        prefill: None,
101        structural_experiment: None,
102        applied_structural_experiment: None,
103    };
104
105    let (delta_tx, mut delta_rx) = tokio::sync::mpsc::unbounded_channel::<String>();
106    let started = Instant::now();
107    let first_delta = tokio::spawn(async move { delta_rx.recv().await.map(|_| started.elapsed()) });
108    let result = vm_call_llm_full_streaming(&opts, delta_tx)
109        .await
110        .map_err(vm_error_message);
111    let latency_ms = duration_ms(started.elapsed());
112    let first_token_ms = first_delta.await.ok().flatten().map(duration_ms);
113    let result = result?;
114
115    Ok(ModelSmokeTestResult {
116        model_id: result.model.clone(),
117        provider: result.provider.clone(),
118        latency_ms,
119        first_token_ms,
120        input_tokens: result.input_tokens,
121        output_tokens: result.output_tokens,
122        estimated_cost_usd: super::calculate_cost_for_provider(
123            &result.provider,
124            &result.model,
125            result.input_tokens,
126            result.output_tokens,
127        ),
128    })
129}
130
131fn duration_ms(duration: std::time::Duration) -> u64 {
132    u64::try_from(duration.as_millis()).unwrap_or(u64::MAX)
133}
134
135fn vm_error_message(error: VmError) -> String {
136    match error {
137        VmError::CategorizedError { message, .. } => message,
138        VmError::Thrown(VmValue::String(message)) => message.to_string(),
139        VmError::Thrown(VmValue::Dict(dict)) => dict
140            .get("message")
141            .map(VmValue::display)
142            .unwrap_or_else(|| VmError::Thrown(VmValue::Dict(dict)).to_string()),
143        other => other.to_string(),
144    }
145}
146
147#[cfg(test)]
148mod tests {
149    use super::{run_model_smoke_test, ModelSmokeTestOptions};
150
151    #[tokio::test]
152    async fn mock_provider_smoke_test_reports_timing_tokens_and_cost() {
153        crate::llm::reset_llm_state();
154        let result = run_model_smoke_test(ModelSmokeTestOptions {
155            model: "mock".to_string(),
156            provider: Some("mock".to_string()),
157            prompt: "ping".to_string(),
158        })
159        .await
160        .expect("mock provider smoke test should not require network");
161
162        assert_eq!(result.model_id, "mock");
163        assert_eq!(result.provider, "mock");
164        assert_eq!(result.input_tokens, 4);
165        assert_eq!(result.output_tokens, 30);
166        assert_eq!(result.estimated_cost_usd, 0.0);
167        assert!(result.first_token_ms.is_some());
168    }
169}
harn_vm/llm/model_test.rs

harn_vm/llm/
model_test.rs
