rig_core/providers/mistral/
completion.rs

1use serde::{Deserialize, Serialize};
2use std::{convert::Infallible, str::FromStr};
3use tracing::{Instrument, Level, enabled, info_span};
4
5use super::client::{Client, Usage};
6use crate::completion::GetTokenUsage;
7use crate::http_client::{self, HttpClientExt};
8use crate::providers::internal::buffered;
9use crate::streaming::{RawStreamingChoice, RawStreamingToolCall, StreamingCompletionResponse};
10use crate::{
11    OneOrMany,
12    completion::{self, CompletionError, CompletionRequest},
13    json_utils, message,
14    providers::mistral::client::ApiResponse,
15    telemetry::SpanCombinator,
16};
17
18/// The latest version of the `codestral` Mistral model
19pub const CODESTRAL: &str = "codestral-latest";
20/// The latest version of the `mistral-large` Mistral model
21pub const MISTRAL_LARGE: &str = "mistral-large-latest";
22/// The latest version of the `pixtral-large` Mistral multimodal model
23pub const PIXTRAL_LARGE: &str = "pixtral-large-latest";
24/// The latest version of the `mistral` Mistral multimodal model, trained on datasets from the Middle East & South Asia
25pub const MISTRAL_SABA: &str = "mistral-saba-latest";
26/// The latest version of the `mistral-3b` Mistral completions model
27pub const MINISTRAL_3B: &str = "ministral-3b-latest";
28/// The latest version of the `mistral-8b` Mistral completions model
29pub const MINISTRAL_8B: &str = "ministral-8b-latest";
30
31/// The latest version of the `mistral-small` Mistral completions model
32pub const MISTRAL_SMALL: &str = "mistral-small-latest";
33/// The `24-09` version of the `pixtral-small` Mistral multimodal model
34pub const PIXTRAL_SMALL: &str = "pixtral-12b-2409";
35/// The `open-mistral-nemo` model
36pub const MISTRAL_NEMO: &str = "open-mistral-nemo";
37/// The `open-mistral-mamba` model
38pub const CODESTRAL_MAMBA: &str = "open-codestral-mamba";
39
40// =================================================================
41// Rig Implementation Types
42// =================================================================
43
44#[derive(Debug, Serialize, Deserialize, PartialEq, Clone)]
45#[serde(tag = "type", rename_all = "lowercase")]
46pub struct AssistantContent {
47    text: String,
48}
49
50#[derive(Debug, Serialize, Deserialize, PartialEq, Clone)]
51#[serde(tag = "type", rename_all = "lowercase")]
52pub enum UserContent {
53    Text { text: String },
54}
55
56#[derive(Debug, Serialize, Deserialize, Clone)]
57pub struct Choice {
58    pub index: usize,
59    pub message: Message,
60    pub logprobs: Option<serde_json::Value>,
61    pub finish_reason: String,
62}
63
64#[derive(Debug, Serialize, Deserialize, PartialEq, Clone)]
65#[serde(tag = "role", rename_all = "lowercase")]
66pub enum Message {
67    User {
68        content: String,
69    },
70    Assistant {
71        content: String,
72        #[serde(
73            default,
74            deserialize_with = "json_utils::null_or_vec",
75            skip_serializing_if = "Vec::is_empty"
76        )]
77        tool_calls: Vec<ToolCall>,
78        #[serde(default)]
79        prefix: bool,
80    },
81    System {
82        content: String,
83    },
84    Tool {
85        /// The name of the tool that was called
86        name: String,
87        /// The content of the tool call
88        content: String,
89        /// The id of the tool call
90        tool_call_id: String,
91    },
92}
93
94impl Message {
95    pub fn user(content: String) -> Self {
96        Message::User { content }
97    }
98
99    pub fn assistant(content: String, tool_calls: Vec<ToolCall>, prefix: bool) -> Self {
100        Message::Assistant {
101            content,
102            tool_calls,
103            prefix,
104        }
105    }
106
107    pub fn system(content: String) -> Self {
108        Message::System { content }
109    }
110}
111
112impl TryFrom<message::Message> for Vec<Message> {
113    type Error = message::MessageError;
114
115    fn try_from(message: message::Message) -> Result<Self, Self::Error> {
116        match message {
117            message::Message::System { content } => Ok(vec![Message::System { content }]),
118            message::Message::User { content } => {
119                let mut tool_result_messages = Vec::new();
120                let mut other_messages = Vec::new();
121
122                for content_item in content {
123                    match content_item {
124                        message::UserContent::ToolResult(message::ToolResult {
125                            id,
126                            call_id,
127                            content: tool_content,
128                        }) => {
129                            let call_id_key = call_id.unwrap_or_else(|| id.clone());
130                            let content_text = tool_content
131                                .into_iter()
132                                .find_map(|content_item| match content_item {
133                                    message::ToolResultContent::Text(text) => Some(text.text),
134                                    message::ToolResultContent::Image(_) => None,
135                                })
136                                .unwrap_or_default();
137                            tool_result_messages.push(Message::Tool {
138                                name: id,
139                                content: content_text,
140                                tool_call_id: call_id_key,
141                            });
142                        }
143                        message::UserContent::Text(message::Text { text, .. }) => {
144                            other_messages.push(Message::User { content: text });
145                        }
146                        _ => {}
147                    }
148                }
149
150                tool_result_messages.append(&mut other_messages);
151                Ok(tool_result_messages)
152            }
153            message::Message::Assistant { content, .. } => {
154                let mut text_content = Vec::new();
155                let mut tool_calls = Vec::new();
156
157                for content in content {
158                    match content {
159                        message::AssistantContent::Text(text) => text_content.push(text),
160                        message::AssistantContent::ToolCall(tool_call) => {
161                            tool_calls.push(tool_call)
162                        }
163                        message::AssistantContent::Reasoning(_) => {
164                            // Mistral conversion path currently does not support assistant-history
165                            // reasoning items. Silently skip to avoid crashing the process.
166                        }
167                        message::AssistantContent::Image(_) => {
168                            return Err(message::MessageError::ConversionError(
169                                "Mistral assistant messages do not support image content".into(),
170                            ));
171                        }
172                    }
173                }
174
175                if text_content.is_empty() && tool_calls.is_empty() {
176                    return Ok(vec![]);
177                }
178
179                Ok(vec![Message::Assistant {
180                    content: text_content
181                        .into_iter()
182                        .next()
183                        .map(|content| content.text)
184                        .unwrap_or_default(),
185                    tool_calls: tool_calls
186                        .into_iter()
187                        .map(|tool_call| tool_call.into())
188                        .collect::<Vec<_>>(),
189                    prefix: false,
190                }])
191            }
192        }
193    }
194}
195
196#[derive(Debug, Serialize, Deserialize, PartialEq, Clone)]
197pub struct ToolCall {
198    pub id: String,
199    #[serde(default)]
200    pub r#type: ToolType,
201    pub function: Function,
202}
203
204impl From<message::ToolCall> for ToolCall {
205    fn from(tool_call: message::ToolCall) -> Self {
206        Self {
207            id: tool_call.id,
208            r#type: ToolType::default(),
209            function: Function {
210                name: tool_call.function.name,
211                arguments: tool_call.function.arguments,
212            },
213        }
214    }
215}
216
217#[derive(Debug, Serialize, Deserialize, PartialEq, Clone)]
218pub struct Function {
219    pub name: String,
220    #[serde(with = "json_utils::stringified_json")]
221    pub arguments: serde_json::Value,
222}
223
224#[derive(Default, Debug, Serialize, Deserialize, PartialEq, Clone)]
225#[serde(rename_all = "lowercase")]
226pub enum ToolType {
227    #[default]
228    Function,
229}
230
231#[derive(Debug, Deserialize, Serialize, Clone)]
232pub struct ToolDefinition {
233    pub r#type: String,
234    pub function: completion::ToolDefinition,
235}
236
237impl From<completion::ToolDefinition> for ToolDefinition {
238    fn from(tool: completion::ToolDefinition) -> Self {
239        Self {
240            r#type: "function".into(),
241            function: tool,
242        }
243    }
244}
245
246#[derive(Debug, Serialize, Deserialize, PartialEq, Clone)]
247pub struct ToolResultContent {
248    #[serde(default)]
249    r#type: ToolResultContentType,
250    text: String,
251}
252
253#[derive(Default, Debug, Serialize, Deserialize, PartialEq, Clone)]
254#[serde(rename_all = "lowercase")]
255pub enum ToolResultContentType {
256    #[default]
257    Text,
258}
259
260impl From<String> for ToolResultContent {
261    fn from(s: String) -> Self {
262        ToolResultContent {
263            r#type: ToolResultContentType::default(),
264            text: s,
265        }
266    }
267}
268
269impl From<String> for UserContent {
270    fn from(s: String) -> Self {
271        UserContent::Text { text: s }
272    }
273}
274
275impl FromStr for UserContent {
276    type Err = Infallible;
277
278    fn from_str(s: &str) -> Result<Self, Self::Err> {
279        Ok(UserContent::Text {
280            text: s.to_string(),
281        })
282    }
283}
284
285impl From<String> for AssistantContent {
286    fn from(s: String) -> Self {
287        AssistantContent { text: s }
288    }
289}
290
291impl FromStr for AssistantContent {
292    type Err = Infallible;
293
294    fn from_str(s: &str) -> Result<Self, Self::Err> {
295        Ok(AssistantContent {
296            text: s.to_string(),
297        })
298    }
299}
300
301#[derive(Clone)]
302pub struct CompletionModel<T = reqwest::Client> {
303    pub(crate) client: Client<T>,
304    pub model: String,
305}
306
307#[derive(Debug, Default, Serialize, Deserialize)]
308pub enum ToolChoice {
309    #[default]
310    Auto,
311    None,
312    Any,
313}
314
315impl TryFrom<message::ToolChoice> for ToolChoice {
316    type Error = CompletionError;
317
318    fn try_from(value: message::ToolChoice) -> Result<Self, Self::Error> {
319        let res = match value {
320            message::ToolChoice::Auto => Self::Auto,
321            message::ToolChoice::None => Self::None,
322            message::ToolChoice::Required => Self::Any,
323            message::ToolChoice::Specific { .. } => {
324                return Err(CompletionError::ProviderError(
325                    "Mistral doesn't support requiring specific tools to be called".to_string(),
326                ));
327            }
328        };
329
330        Ok(res)
331    }
332}
333
334#[derive(Debug, Serialize, Deserialize)]
335pub(super) struct MistralCompletionRequest {
336    model: String,
337    pub messages: Vec<Message>,
338    #[serde(skip_serializing_if = "Option::is_none")]
339    temperature: Option<f64>,
340    #[serde(skip_serializing_if = "Vec::is_empty")]
341    tools: Vec<ToolDefinition>,
342    #[serde(skip_serializing_if = "Option::is_none")]
343    tool_choice: Option<crate::providers::openai::completion::ToolChoice>,
344    #[serde(flatten, skip_serializing_if = "Option::is_none")]
345    pub additional_params: Option<serde_json::Value>,
346}
347
348impl TryFrom<(&str, CompletionRequest)> for MistralCompletionRequest {
349    type Error = CompletionError;
350
351    fn try_from((model, req): (&str, CompletionRequest)) -> Result<Self, Self::Error> {
352        let chat_history = req.chat_history_with_documents();
353        if req.output_schema.is_some() {
354            tracing::warn!("Structured outputs currently not supported for Mistral");
355        }
356        let model = req.model.clone().unwrap_or_else(|| model.to_string());
357        let mut full_history: Vec<Message> = match &req.preamble {
358            Some(preamble) => vec![Message::system(preamble.clone())],
359            None => vec![],
360        };
361        let chat_history: Vec<Message> = chat_history
362            .into_iter()
363            .map(|message| message.try_into())
364            .collect::<Result<Vec<Vec<Message>>, _>>()?
365            .into_iter()
366            .flatten()
367            .collect();
368
369        full_history.extend(chat_history);
370
371        if full_history.is_empty() {
372            return Err(CompletionError::RequestError(
373                std::io::Error::new(
374                    std::io::ErrorKind::InvalidInput,
375                    "Mistral request has no provider-compatible messages after conversion",
376                )
377                .into(),
378            ));
379        }
380
381        let tool_choice = req
382            .tool_choice
383            .clone()
384            .map(crate::providers::openai::completion::ToolChoice::try_from)
385            .transpose()?;
386
387        Ok(Self {
388            model: model.to_string(),
389            messages: full_history,
390            temperature: req.temperature,
391            tools: req
392                .tools
393                .clone()
394                .into_iter()
395                .map(ToolDefinition::from)
396                .collect::<Vec<_>>(),
397            tool_choice,
398            additional_params: req.additional_params,
399        })
400    }
401}
402
403impl<T> CompletionModel<T> {
404    pub fn new(client: Client<T>, model: impl Into<String>) -> Self {
405        Self {
406            client,
407            model: model.into(),
408        }
409    }
410
411    pub fn with_model(client: Client<T>, model: &str) -> Self {
412        Self {
413            client,
414            model: model.into(),
415        }
416    }
417}
418
419#[derive(Debug, Deserialize, Clone, Serialize)]
420pub struct CompletionResponse {
421    pub id: String,
422    pub object: String,
423    pub created: u64,
424    pub model: String,
425    pub system_fingerprint: Option<String>,
426    pub choices: Vec<Choice>,
427    pub usage: Option<Usage>,
428}
429
430impl crate::telemetry::ProviderResponseExt for CompletionResponse {
431    type OutputMessage = Choice;
432    type Usage = Usage;
433
434    fn get_response_id(&self) -> Option<String> {
435        Some(self.id.clone())
436    }
437
438    fn get_response_model_name(&self) -> Option<String> {
439        Some(self.model.clone())
440    }
441
442    fn get_output_messages(&self) -> Vec<Self::OutputMessage> {
443        self.choices.clone()
444    }
445
446    fn get_text_response(&self) -> Option<String> {
447        let res = self
448            .choices
449            .iter()
450            .filter_map(|choice| match choice.message {
451                Message::Assistant { ref content, .. } => {
452                    if content.is_empty() {
453                        None
454                    } else {
455                        Some(content.to_string())
456                    }
457                }
458                _ => None,
459            })
460            .collect::<Vec<String>>()
461            .join("\n");
462
463        if res.is_empty() { None } else { Some(res) }
464    }
465
466    fn get_usage(&self) -> Option<Self::Usage> {
467        self.usage.clone()
468    }
469}
470
471impl GetTokenUsage for CompletionResponse {
472    fn token_usage(&self) -> crate::completion::Usage {
473        let Some(api_usage) = self.usage.as_ref() else {
474            return crate::completion::Usage::new();
475        };
476
477        let mut usage = crate::completion::Usage::new();
478        usage.input_tokens = api_usage.prompt_tokens as u64;
479        usage.output_tokens = api_usage.completion_tokens as u64;
480        usage.total_tokens = api_usage.total_tokens as u64;
481        usage.cached_input_tokens = api_usage.cached_tokens();
482
483        usage
484    }
485}
486
487impl TryFrom<CompletionResponse> for completion::CompletionResponse<CompletionResponse> {
488    type Error = CompletionError;
489
490    fn try_from(response: CompletionResponse) -> Result<Self, Self::Error> {
491        let choice = response.choices.first().ok_or_else(|| {
492            CompletionError::ResponseError("Response contained no choices".to_owned())
493        })?;
494        let content = match &choice.message {
495            Message::Assistant {
496                content,
497                tool_calls,
498                ..
499            } => {
500                let mut content = if content.is_empty() {
501                    vec![]
502                } else {
503                    vec![completion::AssistantContent::text(content.clone())]
504                };
505
506                content.extend(
507                    tool_calls
508                        .iter()
509                        .map(|call| {
510                            completion::AssistantContent::tool_call(
511                                &call.id,
512                                &call.function.name,
513                                call.function.arguments.clone(),
514                            )
515                        })
516                        .collect::<Vec<_>>(),
517                );
518                Ok(content)
519            }
520            _ => Err(CompletionError::ResponseError(
521                "Response did not contain a valid message or tool call".into(),
522            )),
523        }?;
524
525        let choice = OneOrMany::many(content).map_err(|_| {
526            CompletionError::ResponseError(
527                "Response contained no message or tool call (empty)".to_owned(),
528            )
529        })?;
530
531        let usage = response
532            .usage
533            .as_ref()
534            .map(|usage| completion::Usage {
535                input_tokens: usage.prompt_tokens as u64,
536                output_tokens: (usage.total_tokens - usage.prompt_tokens) as u64,
537                total_tokens: usage.total_tokens as u64,
538                cached_input_tokens: usage.cached_tokens(),
539                cache_creation_input_tokens: 0,
540                tool_use_prompt_tokens: 0,
541                reasoning_tokens: 0,
542            })
543            .unwrap_or_default();
544
545        Ok(completion::CompletionResponse {
546            choice,
547            usage,
548            raw_response: response,
549            message_id: None,
550        })
551    }
552}
553
554fn assistant_content_to_streaming_choices(
555    content: message::AssistantContent,
556) -> Result<Vec<RawStreamingChoice<CompletionResponse>>, CompletionError> {
557    match content {
558        message::AssistantContent::Text(t) => Ok(vec![RawStreamingChoice::Message(t.text)]),
559        message::AssistantContent::ToolCall(tc) => Ok(vec![RawStreamingChoice::ToolCall(
560            RawStreamingToolCall::new(tc.id, tc.function.name, tc.function.arguments),
561        )]),
562        message::AssistantContent::Reasoning(_) => Ok(Vec::new()),
563        message::AssistantContent::Image(_) => Err(CompletionError::ResponseError(
564            "Image content is not supported on Mistral via Rig".into(),
565        )),
566    }
567}
568
569impl<T> completion::CompletionModel for CompletionModel<T>
570where
571    T: HttpClientExt + Send + Clone + std::fmt::Debug + 'static,
572{
573    type Response = CompletionResponse;
574    type StreamingResponse = CompletionResponse;
575
576    type Client = Client<T>;
577
578    fn make(client: &Self::Client, model: impl Into<String>) -> Self {
579        Self::new(client.clone(), model.into())
580    }
581
582    async fn completion(
583        &self,
584        completion_request: CompletionRequest,
585    ) -> Result<completion::CompletionResponse<CompletionResponse>, CompletionError> {
586        let preamble = completion_request.preamble.clone();
587        let request =
588            MistralCompletionRequest::try_from((self.model.as_ref(), completion_request))?;
589
590        if enabled!(Level::TRACE) {
591            tracing::trace!(
592                target: "rig::completions",
593                "Mistral completion request: {}",
594                serde_json::to_string_pretty(&request)?
595            );
596        }
597
598        let span = if tracing::Span::current().is_disabled() {
599            info_span!(
600                target: "rig::completions",
601                "chat",
602                gen_ai.operation.name = "chat",
603                gen_ai.provider.name = "mistral",
604                gen_ai.request.model = self.model,
605                gen_ai.system_instructions = &preamble,
606                gen_ai.response.id = tracing::field::Empty,
607                gen_ai.response.model = tracing::field::Empty,
608                gen_ai.usage.output_tokens = tracing::field::Empty,
609                gen_ai.usage.input_tokens = tracing::field::Empty,
610                gen_ai.usage.cache_read.input_tokens = tracing::field::Empty,
611            )
612        } else {
613            tracing::Span::current()
614        };
615
616        let body = serde_json::to_vec(&request)?;
617
618        let request = self
619            .client
620            .post("v1/chat/completions")?
621            .body(body)
622            .map_err(|e| CompletionError::HttpError(e.into()))?;
623
624        async move {
625            let response = self.client.send(request).await?;
626
627            if response.status().is_success() {
628                let text = http_client::text(response).await?;
629                match serde_json::from_str::<ApiResponse<CompletionResponse>>(&text)? {
630                    ApiResponse::Ok(response) => {
631                        let span = tracing::Span::current();
632                        span.record_token_usage(&response);
633                        span.record_response_metadata(&response);
634                        response.try_into()
635                    }
636                    ApiResponse::Err(err) => Err(CompletionError::ProviderError(err.message)),
637                }
638            } else {
639                let text = http_client::text(response).await?;
640                Err(CompletionError::ProviderError(text))
641            }
642        }
643        .instrument(span)
644        .await
645    }
646
647    async fn stream(
648        &self,
649        request: CompletionRequest,
650    ) -> Result<StreamingCompletionResponse<Self::StreamingResponse>, CompletionError> {
651        let resp = self.completion(request).await?;
652        buffered::stream_from_completion_response(resp, assistant_content_to_streaming_choices)
653    }
654}
655
656#[cfg(test)]
657mod tests {
658    use super::*;
659
660    #[test]
661    fn test_response_deserialization() {
662        //https://docs.mistral.ai/api/#tag/chat/operation/chat_completion_v1_chat_completions_post
663        let json_data = r#"
664        {
665            "id": "cmpl-e5cc70bb28c444948073e77776eb30ef",
666            "object": "chat.completion",
667            "model": "mistral-small-latest",
668            "usage": {
669                "prompt_tokens": 16,
670                "completion_tokens": 34,
671                "total_tokens": 50
672            },
673            "created": 1702256327,
674            "choices": [
675                {
676                    "index": 0,
677                    "message": {
678                        "content": "string",
679                        "tool_calls": [
680                            {
681                                "id": "null",
682                                "type": "function",
683                                "function": {
684                                    "name": "string",
685                                    "arguments": "{ }"
686                                },
687                                "index": 0
688                            }
689                        ],
690                        "prefix": false,
691                        "role": "assistant"
692                    },
693                    "finish_reason": "stop"
694                }
695            ]
696        }
697        "#;
698        let completion_response = serde_json::from_str::<CompletionResponse>(json_data).unwrap();
699        assert_eq!(completion_response.model, MISTRAL_SMALL);
700
701        let CompletionResponse {
702            id,
703            object,
704            created,
705            choices,
706            usage,
707            ..
708        } = completion_response;
709
710        assert_eq!(id, "cmpl-e5cc70bb28c444948073e77776eb30ef");
711
712        let usage = usage.unwrap();
713        assert_eq!(usage.prompt_tokens, 16);
714        assert_eq!(usage.completion_tokens, 34);
715        assert_eq!(usage.total_tokens, 50);
716        assert_eq!(usage.cached_tokens(), 0);
717        assert!(usage.prompt_tokens_details.is_none());
718        assert!(usage.num_cached_tokens.is_none());
719        assert_eq!(object, "chat.completion".to_string());
720        assert_eq!(created, 1702256327);
721        assert_eq!(choices.len(), 1);
722    }
723
724    #[test]
725    fn test_usage_deserializes_prompt_tokens_details_cached_tokens() {
726        let json = r#"{
727            "prompt_tokens": 100,
728            "completion_tokens": 20,
729            "total_tokens": 120,
730            "prompt_tokens_details": { "cached_tokens": 42 }
731        }"#;
732        let usage: Usage = serde_json::from_str(json).unwrap();
733        assert_eq!(usage.prompt_tokens, 100);
734        assert_eq!(
735            usage.prompt_tokens_details.as_ref().unwrap().cached_tokens,
736            42
737        );
738        assert_eq!(usage.cached_tokens(), 42);
739    }
740
741    #[test]
742    fn test_usage_accepts_singular_prompt_token_details_alias() {
743        let json = r#"{
744            "prompt_tokens": 100,
745            "completion_tokens": 20,
746            "total_tokens": 120,
747            "prompt_token_details": { "cached_tokens": 7 }
748        }"#;
749        let usage: Usage = serde_json::from_str(json).unwrap();
750        assert_eq!(
751            usage.prompt_tokens_details.as_ref().unwrap().cached_tokens,
752            7
753        );
754        assert_eq!(usage.cached_tokens(), 7);
755    }
756
757    #[test]
758    fn test_usage_falls_back_to_num_cached_tokens() {
759        let json = r#"{
760            "prompt_tokens": 100,
761            "completion_tokens": 20,
762            "total_tokens": 120,
763            "num_cached_tokens": 13
764        }"#;
765        let usage: Usage = serde_json::from_str(json).unwrap();
766        assert_eq!(usage.num_cached_tokens, Some(13));
767        assert!(usage.prompt_tokens_details.is_none());
768        assert_eq!(usage.cached_tokens(), 13);
769    }
770
771    #[test]
772    fn test_usage_prefers_prompt_tokens_details_over_num_cached_tokens() {
773        let json = r#"{
774            "prompt_tokens": 100,
775            "completion_tokens": 20,
776            "total_tokens": 120,
777            "num_cached_tokens": 1,
778            "prompt_tokens_details": { "cached_tokens": 99 }
779        }"#;
780        let usage: Usage = serde_json::from_str(json).unwrap();
781        assert_eq!(usage.cached_tokens(), 99);
782    }
783
784    #[test]
785    fn test_token_usage_threads_cached_tokens_into_completion_usage() {
786        let json = r#"{
787            "id": "cmpl-x",
788            "object": "chat.completion",
789            "model": "mistral-small-latest",
790            "created": 1700000000,
791            "choices": [{
792                "index": 0,
793                "message": { "content": "hi", "role": "assistant", "prefix": false },
794                "finish_reason": "stop"
795            }],
796            "usage": {
797                "prompt_tokens": 100,
798                "completion_tokens": 20,
799                "total_tokens": 120,
800                "prompt_tokens_details": { "cached_tokens": 42 }
801            }
802        }"#;
803        let response: CompletionResponse = serde_json::from_str(json).unwrap();
804        let usage = response.token_usage();
805        assert_eq!(usage.input_tokens, 100);
806        assert_eq!(usage.output_tokens, 20);
807        assert_eq!(usage.total_tokens, 120);
808        assert_eq!(usage.cached_input_tokens, 42);
809    }
810
811    #[test]
812    fn test_assistant_reasoning_is_skipped_in_message_conversion() {
813        let assistant = message::Message::Assistant {
814            id: None,
815            content: OneOrMany::one(message::AssistantContent::reasoning("hidden")),
816        };
817
818        let converted: Vec<Message> = assistant.try_into().expect("conversion should work");
819        assert!(converted.is_empty());
820    }
821
822    #[test]
823    fn test_assistant_text_and_tool_call_are_preserved_when_reasoning_present() {
824        let assistant = message::Message::Assistant {
825            id: None,
826            content: OneOrMany::many(vec![
827                message::AssistantContent::reasoning("hidden"),
828                message::AssistantContent::text("visible"),
829                message::AssistantContent::tool_call(
830                    "call_1",
831                    "subtract",
832                    serde_json::json!({"x": 2, "y": 1}),
833                ),
834            ])
835            .expect("non-empty assistant content"),
836        };
837
838        let converted: Vec<Message> = assistant.try_into().expect("conversion should work");
839        assert_eq!(converted.len(), 1);
840
841        match &converted[0] {
842            Message::Assistant {
843                content,
844                tool_calls,
845                ..
846            } => {
847                assert_eq!(content, "visible");
848                assert_eq!(tool_calls.len(), 1);
849                assert_eq!(tool_calls[0].id, "call_1");
850                assert_eq!(tool_calls[0].function.name, "subtract");
851                assert_eq!(
852                    tool_calls[0].function.arguments,
853                    serde_json::json!({"x": 2, "y": 1})
854                );
855            }
856            _ => panic!("expected assistant message"),
857        }
858    }
859
860    #[test]
861    fn test_streaming_choice_mapping_skips_reasoning_and_preserves_other_content() {
862        let reasoning_choices =
863            assistant_content_to_streaming_choices(message::AssistantContent::reasoning("hidden"))
864                .expect("reasoning should be ignored");
865        assert!(reasoning_choices.is_empty());
866
867        let text_choices =
868            assistant_content_to_streaming_choices(message::AssistantContent::text("visible"))
869                .expect("text should be preserved");
870        match text_choices.as_slice() {
871            [RawStreamingChoice::Message(text)] => assert_eq!(text, "visible"),
872            _ => panic!("expected text streaming choice"),
873        }
874
875        let tool_choices =
876            assistant_content_to_streaming_choices(message::AssistantContent::tool_call(
877                "call_2",
878                "add",
879                serde_json::json!({"x": 2, "y": 3}),
880            ))
881            .expect("tool call should be preserved");
882        match tool_choices.as_slice() {
883            [RawStreamingChoice::ToolCall(call)] => {
884                assert_eq!(call.id, "call_2");
885                assert_eq!(call.name, "add");
886                assert_eq!(call.arguments, serde_json::json!({"x": 2, "y": 3}));
887            }
888            _ => panic!("expected tool-call streaming choice"),
889        }
890    }
891
892    #[test]
893    fn test_request_conversion_errors_when_all_messages_are_filtered() {
894        let request = CompletionRequest {
895            preamble: None,
896            chat_history: OneOrMany::one(message::Message::Assistant {
897                id: None,
898                content: OneOrMany::one(message::AssistantContent::reasoning("hidden")),
899            }),
900            documents: vec![],
901            tools: vec![],
902            temperature: None,
903            max_tokens: None,
904            tool_choice: None,
905            additional_params: None,
906            model: None,
907            output_schema: None,
908        };
909
910        let result = MistralCompletionRequest::try_from((MISTRAL_SMALL, request));
911        assert!(matches!(result, Err(CompletionError::RequestError(_))));
912    }
913}
rig_core/providers/mistral/completion.rs

rig_core/providers/mistral/
completion.rs
