diff --git a/DESCRIPTION b/DESCRIPTION
index 504841e22..f7da45d82 100644
--- a/DESCRIPTION
+++ b/DESCRIPTION
@@ -86,6 +86,7 @@ Collate:
     'import-standalone-types-check.R'
     'interpolate.R'
     'live.R'
+    'model.R'
     'parallel-chat.R'
     'params.R'
     'provider-anthropic.R'
diff --git a/NAMESPACE b/NAMESPACE
index 8750d8f34..8a470a7ac 100644
--- a/NAMESPACE
+++ b/NAMESPACE
@@ -16,6 +16,7 @@ export(ContentText)
 export(ContentThinking)
 export(ContentToolRequest)
 export(ContentToolResult)
+export(Model)
 export(Provider)
 export(SystemTurn)
 export(Turn)
diff --git a/R/batch-chat.R b/R/batch-chat.R
index 0f947421a..f5019ef12 100644
--- a/R/batch-chat.R
+++ b/R/batch-chat.R
@@ -193,6 +193,7 @@ BatchJob <- R6::R6Class(
 
     # Internal state
     provider = NULL,
+    model = NULL,
     started_at = NULL,
     stage = NULL,
     batch = NULL,
@@ -208,6 +209,7 @@ BatchJob <- R6::R6Class(
       call = caller_env(2)
     ) {
       self$provider <- chat$get_provider()
+      self$model <- chat$get_model_obj()
       check_has_batch_support(self$provider, call = call)
 
       user_turns <- as_user_turns(prompts, call = call)
@@ -326,7 +328,7 @@ BatchJob <- R6::R6Class(
 
     retrieve = function() {
       self$results <- batch_retrieve(self$provider, self$batch)
-      log_turns(self$provider, self$result_turns())
+      log_turns(self$provider, self$model, self$result_turns())
 
       self$stage <- "done"
       self$save_state()
@@ -335,7 +337,12 @@ BatchJob <- R6::R6Class(
 
     result_turns = function() {
       map2(self$results, self$user_turns, function(result, user_turn) {
-        batch_result_turn(self$provider, result, has_type = !is.null(self$type))
+        batch_result_turn(
+          self$provider,
+          self$model,
+          result,
+          has_type = !is.null(self$type)
+        )
       })
     },
 
@@ -379,7 +386,6 @@ BatchJob <- R6::R6Class(
 provider_hash <- function(x) {
   list(
     name = x@name,
-    model = x@model,
     base_url = x@base_url
   )
 }
diff --git a/R/chat.R b/R/chat.R
index 0e7da10ba..a863d037b 100644
--- a/R/chat.R
+++ b/R/chat.R
@@ -24,6 +24,7 @@ Chat <- R6::R6Class(
   "Chat",
   public = list(
     #' @param provider A provider object.
+    #' @param model A model object.
     #' @param system_prompt System prompt to start the conversation with.
     #' @param echo One of the following options:
     #'   * `none`: don't emit any output (default when running in a function).
@@ -33,8 +34,14 @@ Chat <- R6::R6Class(
     #'
     #'  Note this only affects the `chat()` method. You can override the default
     #'  by setting the `ellmer_echo` option.
-    initialize = function(provider, system_prompt = NULL, echo = "none") {
+    initialize = function(
+      provider,
+      model,
+      system_prompt = NULL,
+      echo = "none"
+    ) {
       private$provider <- provider
+      private$model <- model
       private$echo <- echo
       private$callback_on_tool_request <- CallbackManager$new(args = "request")
       private$callback_on_tool_result <- CallbackManager$new(args = "result")
@@ -78,7 +85,7 @@ Chat <- R6::R6Class(
       check_turn(assistant)
 
       if (log_tokens) {
-        log_turn(private$provider, assistant)
+        log_turn(private$provider, private$model, assistant)
       }
 
       private$.turns[[length(private$.turns) + 1]] <- user
@@ -97,7 +104,7 @@ Chat <- R6::R6Class(
 
     #' @description Retrieve the model name
     get_model = function() {
-      private$provider@model
+      private$model@name
     },
 
     #' @description Update the system prompt
@@ -381,6 +388,11 @@ Chat <- R6::R6Class(
       private$provider
     },
 
+    #' @description Get the underlying model object. For expert use only.
+    get_model_obj = function() {
+      private$model
+    },
+
     #' @description Retrieve the list of registered tools.
     get_tools = function() {
       private$tools
@@ -422,6 +434,7 @@ Chat <- R6::R6Class(
   ),
   private = list(
     provider = NULL,
+    model = NULL,
 
     .turns = list(),
     echo = NULL,
@@ -579,6 +592,7 @@ Chat <- R6::R6Class(
 
       response <- chat_perform(
         provider = private$provider,
+        model = private$model,
         mode = if (stream) "stream" else "value",
         turns = c(private$.turns, list(user_turn)),
         tools = if (is.null(type)) private$tools,
@@ -603,11 +617,17 @@ Chat <- R6::R6Class(
 
           result <- stream_merge_chunks(private$provider, result, chunk)
         }
-        turn <- value_turn(private$provider, result, has_type = !is.null(type))
+        turn <- value_turn(
+          private$provider,
+          private$model,
+          result,
+          has_type = !is.null(type)
+        )
         turn <- match_tools(turn, private$tools)
       } else {
         turn <- value_turn(
           private$provider,
+          private$model,
           resp_body_json(response),
           has_type = !is.null(type)
         )
@@ -661,6 +681,7 @@ Chat <- R6::R6Class(
     ) {
       response <- chat_perform(
         provider = private$provider,
+        model = private$model,
         mode = if (stream) "async-stream" else "async-value",
         turns = c(private$.turns, list(user_turn)),
         tools = if (is.null(type)) private$tools,
@@ -685,12 +706,18 @@ Chat <- R6::R6Class(
 
           result <- stream_merge_chunks(private$provider, result, chunk)
         }
-        turn <- value_turn(private$provider, result, has_type = !is.null(type))
+        turn <- value_turn(
+          private$provider,
+          private$model,
+          result,
+          has_type = !is.null(type)
+        )
       } else {
         result <- await(response)
 
         turn <- value_turn(
           private$provider,
+          private$model,
           resp_body_json(result),
           has_type = !is.null(type)
         )
@@ -738,6 +765,7 @@ Chat <- R6::R6Class(
 #' @export
 print.Chat <- function(x, ...) {
   provider <- x$get_provider()
+  model <- x$get_model()
   turns <- x$get_turns(include_system_prompt = TRUE)
 
   assistant_turns <- keep(turns, \(x) x@role == "assistant")
@@ -746,7 +774,7 @@ print.Chat <- function(x, ...) {
 
   cat(paste_c(
     "<Chat",
-    c(" ", provider@name, "/", provider@model),
+    c(" ", provider@name, "/", model),
     c(" turns=", length(turns)),
     turn_cost(total_tokens, total_cost, prefix = " "),
     ">\n"
diff --git a/R/httr2.R b/R/httr2.R
index 7a20e2839..1613b7747 100644
--- a/R/httr2.R
+++ b/R/httr2.R
@@ -3,6 +3,7 @@
 # We will recconsider this in the future if necessary.
 chat_perform <- function(
   provider,
+  model,
   mode = c("value", "stream", "async-stream", "async-value"),
   turns,
   tools = NULL,
@@ -14,6 +15,7 @@ chat_perform <- function(
 
   req <- chat_request(
     provider = provider,
+    model = model,
     turns = turns,
     tools = tools,
     stream = stream,
diff --git a/R/model.R b/R/model.R
new file mode 100644
index 000000000..56c5dde8b
--- /dev/null
+++ b/R/model.R
@@ -0,0 +1,29 @@
+#' A chatbot model
+#'
+#' A Model captures the details of a specific language model and its
+#' configuration parameters. This includes the model name, parameters like
+#' temperature and max_tokens, and any extra arguments to include in API
+#' requests.
+#'
+#' Model objects are typically created internally by [chat_openai()],
+#' [chat_anthropic()], and other `chat_*()` functions. You generally won't
+#' need to create them directly unless you're implementing a custom provider.
+#'
+#' @export
+#' @param name Name of the model (e.g., "gpt-4", "claude-sonnet-4").
+#' @param params A list of standard parameters created by [params()].
+#' @param extra_args Arbitrary extra arguments to be included in the request body.
+#' @return An S7 Model object.
+#' @examples
+#' Model(
+#'   name = "gpt-4",
+#'   params = params(temperature = 0.7)
+#' )
+Model <- new_class(
+  "Model",
+  properties = list(
+    name = prop_string(),
+    params = class_list,
+    extra_args = class_list
+  )
+)
diff --git a/R/parallel-chat.R b/R/parallel-chat.R
index 3b232f677..f927896b0 100644
--- a/R/parallel-chat.R
+++ b/R/parallel-chat.R
@@ -78,6 +78,7 @@ parallel_chat <- function(
   my_parallel_turns <- function(conversations) {
     parallel_turns(
       provider = chat$get_provider(),
+      model = chat$get_model_obj(),
       conversations = conversations,
       tools = chat$get_tools(),
       max_active = max_active,
@@ -127,7 +128,7 @@ parallel_chat <- function(
   map(seq_along(conversations), function(i) {
     if (is_ok[[i]]) {
       turns <- conversations[[i]]
-      log_turns(chat$get_provider(), turns)
+      log_turns(chat$get_provider(), chat$get_model_obj(), turns)
       chat$clone()$set_turns(turns)
     } else {
       assistant_turns[[i]]
@@ -191,6 +192,7 @@ parallel_chat_structured <- function(
   on_error <- arg_match(on_error)
 
   provider <- chat$get_provider()
+  model <- chat$get_model_obj()
   needs_wrapper <- type_needs_wrapper(type, provider)
 
   # First build up list of cumulative conversations
@@ -200,6 +202,7 @@ parallel_chat_structured <- function(
 
   turns <- parallel_turns(
     provider = provider,
+    model = model,
     conversations = conversations,
     tools = chat$get_tools(),
     type = wrap_type_if_needed(type, needs_wrapper),
@@ -207,7 +210,7 @@ parallel_chat_structured <- function(
     rpm = rpm,
     on_error = on_error
   )
-  log_turns(provider, turns)
+  log_turns(provider, model, turns)
 
   multi_convert(
     provider,
@@ -227,7 +230,7 @@ multi_convert <- function(
   include_tokens = FALSE,
   include_cost = FALSE
 ) {
-  needs_wrapper <- type_needs_wrapper(type, provider)
+  needs_wrapper <- type_needs_wrapper(type, provider, NULL)
 
   rows <- map(turns, \(turn) {
     if (turn_failed(turn)) {
@@ -306,6 +309,7 @@ turn_failed <- function(turn) {
 
 parallel_turns <- function(
   provider,
+  model,
   conversations,
   tools,
   type = NULL,
@@ -316,6 +320,7 @@ parallel_turns <- function(
   reqs <- map(conversations, function(turns) {
     chat_request(
       provider = provider,
+      model = model,
       turns = turns,
       type = type,
       tools = tools,
@@ -352,7 +357,7 @@ parallel_turns <- function(
       resp
     } else {
       json <- resp_body_json(resp)
-      turn <- value_turn(provider, json, has_type = !is.null(type))
+      turn <- value_turn(provider, model, json, has_type = !is.null(type))
       turn@duration <- resp_timing(resp)[["total"]] %||% NA_real_
       turn
     }
diff --git a/R/provider-anthropic.R b/R/provider-anthropic.R
index b0ae9f4fe..be0711d5e 100644
--- a/R/provider-anthropic.R
+++ b/R/provider-anthropic.R
@@ -100,17 +100,24 @@ chat_anthropic <- function(
 
   provider <- ProviderAnthropic(
     name = "Anthropic",
-    model = model,
-    params = params %||% params(),
-    extra_args = api_args,
     extra_headers = api_headers,
     base_url = base_url,
     beta_headers = beta_headers,
     credentials = credentials,
     cache = cache
   )
+  model_obj <- Model(
+    name = model,
+    params = params %||% params(),
+    extra_args = api_args
+  )
 
-  Chat$new(provider = provider, system_prompt = system_prompt, echo = echo)
+  Chat$new(
+    provider = provider,
+    model = model_obj,
+    system_prompt = system_prompt,
+    echo = echo
+  )
 }
 
 #' @rdname chat_anthropic
@@ -180,6 +187,7 @@ method(chat_path, ProviderAnthropic) <- function(provider) {
 }
 method(chat_body, ProviderAnthropic) <- function(
   provider,
+  model,
   stream = TRUE,
   turns = list(),
   tools = list(),
@@ -213,7 +221,7 @@ method(chat_body, ProviderAnthropic) <- function(
   }
   tools <- as_json(provider, unname(tools))
 
-  params <- chat_params(provider, provider@params)
+  params <- chat_params(provider, model)
   if (has_name(params, "budget_tokens")) {
     thinking <- list(
       type = "enabled",
@@ -225,7 +233,7 @@ method(chat_body, ProviderAnthropic) <- function(
   }
 
   compact(list2(
-    model = provider@model,
+    model = model@name,
     system = system,
     messages = messages,
     stream = stream,
@@ -236,9 +244,9 @@ method(chat_body, ProviderAnthropic) <- function(
   ))
 }
 
-method(chat_params, ProviderAnthropic) <- function(provider, params) {
+method(chat_params, ProviderAnthropic) <- function(provider, model) {
   params <- standardise_params(
-    params,
+    model@params,
     c(
       temperature = "temperature",
       top_p = "top_p",
@@ -326,7 +334,7 @@ method(stream_merge_chunks, ProviderAnthropic) <- function(
   result
 }
 
-method(value_tokens, ProviderAnthropic) <- function(provider, json) {
+method(value_tokens, ProviderAnthropic) <- function(provider, model, json) {
   tokens(
     # Hack in pricing for cache writes
     input = json$usage$input_tokens +
@@ -338,6 +346,7 @@ method(value_tokens, ProviderAnthropic) <- function(provider, json) {
 
 method(value_turn, ProviderAnthropic) <- function(
   provider,
+  model,
   result,
   has_type = FALSE
 ) {
@@ -366,8 +375,8 @@ method(value_turn, ProviderAnthropic) <- function(
     }
   })
 
-  tokens <- value_tokens(provider, result)
-  cost <- get_token_cost(provider, tokens)
+  tokens <- value_tokens(provider, model, result)
+  cost <- get_token_cost(provider, model, tokens)
   AssistantTurn(contents, json = result, tokens = unlist(tokens), cost = cost)
 }
 
@@ -584,11 +593,12 @@ method(batch_retrieve, ProviderAnthropic) <- function(provider, batch) {
 
 method(batch_result_turn, ProviderAnthropic) <- function(
   provider,
+  model,
   result,
   has_type = FALSE
 ) {
   if (result$type == "succeeded") {
-    value_turn(provider, result$message, has_type = has_type)
+    value_turn(provider, model, result$message, has_type = has_type)
   } else {
     NULL
   }
@@ -604,7 +614,6 @@ models_anthropic <- function(
 ) {
   provider <- ProviderAnthropic(
     name = "Anthropic",
-    model = "",
     base_url = base_url,
     credentials = function() api_key,
     cache = "none"
diff --git a/R/provider-aws.R b/R/provider-aws.R
index 0132a13ba..7f1e1bb90 100644
--- a/R/provider-aws.R
+++ b/R/provider-aws.R
@@ -73,11 +73,19 @@ chat_aws_bedrock <- function(
     base_url = base_url,
     model = model,
     profile = profile,
-    params = params,
-    extra_args = api_args,
     extra_headers = api_headers
   )
-  Chat$new(provider = provider, system_prompt = system_prompt, echo = echo)
+  model_obj <- Model(
+    name = model,
+    params = params,
+    extra_args = api_args
+  )
+  Chat$new(
+    provider = provider,
+    model = model_obj,
+    system_prompt = system_prompt,
+    echo = echo
+  )
 }
 
 
@@ -90,7 +98,7 @@ models_aws_bedrock <- function(profile = NULL, base_url = NULL) {
   provider <- provider_aws_bedrock(
     base_url = base_url,
     model = "",
-    profile = profile,
+    profile = profile
   )
   req <- base_request(provider)
   req <- req_url_path_append(req, "foundation-models")
@@ -112,8 +120,6 @@ provider_aws_bedrock <- function(
   base_url,
   model = "",
   profile = NULL,
-  params = list(),
-  extra_args = list(),
   extra_headers = character()
 ) {
   cache <- aws_creds_cache(profile)
@@ -128,13 +134,11 @@ provider_aws_bedrock <- function(
   ProviderAWSBedrock(
     name = "AWS/Bedrock",
     base_url = base_url,
-    model = model,
     profile = profile,
     region = credentials$region,
     cache = cache,
-    params = params,
-    extra_args = extra_args,
-    extra_headers = extra_headers
+    extra_headers = extra_headers,
+    model = model
   )
 }
 
@@ -171,10 +175,10 @@ method(base_request_error, ProviderAWSBedrock) <- function(provider, req) {
   })
 }
 
-method(chat_params, ProviderAWSBedrock) <- function(provider, params) {
+method(chat_params, ProviderAWSBedrock) <- function(provider, model) {
   # https://docs.aws.amazon.com/bedrock/latest/APIReference/API_runtime_InferenceConfiguration.html
   standardise_params(
-    params,
+    model@params,
     c(
       temperature = "temperature",
       topP = "top_p",
@@ -186,6 +190,7 @@ method(chat_params, ProviderAWSBedrock) <- function(provider, params) {
 
 method(chat_request, ProviderAWSBedrock) <- function(
   provider,
+  model,
   stream = TRUE,
   turns = list(),
   tools = list(),
@@ -229,9 +234,9 @@ method(chat_request, ProviderAWSBedrock) <- function(
   }
 
   # Merge params into inferenceConfig, giving precedence to manual api_args
-  params <- chat_params(provider, provider@params)
+  params <- chat_params(provider, model)
 
-  extra_args <- provider@extra_args
+  extra_args <- model@extra_args
   extra_args$inferenceConfig <- modify_list(params, extra_args$inferenceConfig)
 
   # https://docs.aws.amazon.com/bedrock/latest/APIReference/API_runtime_Converse.html
@@ -324,7 +329,7 @@ method(stream_merge_chunks, ProviderAWSBedrock) <- function(
   result
 }
 
-method(value_tokens, ProviderAWSBedrock) <- function(provider, json) {
+method(value_tokens, ProviderAWSBedrock) <- function(provider, model, json) {
   tokens(
     input = json$usage$inputTokens,
     output = json$usage$outputTokens,
@@ -333,6 +338,7 @@ method(value_tokens, ProviderAWSBedrock) <- function(provider, json) {
 
 method(value_turn, ProviderAWSBedrock) <- function(
   provider,
+  model,
   result,
   has_type = FALSE
 ) {
@@ -357,8 +363,8 @@ method(value_turn, ProviderAWSBedrock) <- function(
     }
   })
 
-  tokens <- value_tokens(provider, result)
-  cost <- get_token_cost(provider, tokens)
+  tokens <- value_tokens(provider, model, result)
+  cost <- get_token_cost(provider, model, tokens)
   AssistantTurn(contents, json = result, tokens = unlist(tokens), cost = cost)
 }
 
diff --git a/R/provider-azure.R b/R/provider-azure.R
index 634e3198a..861c64378 100644
--- a/R/provider-azure.R
+++ b/R/provider-azure.R
@@ -77,14 +77,21 @@ chat_azure_openai <- function(
   provider <- ProviderAzureOpenAI(
     name = "Azure/OpenAI",
     base_url = paste0(endpoint, "/openai/deployments/", model),
-    model = model,
-    params = params,
     api_version = api_version,
     credentials = credentials,
-    extra_args = api_args,
     extra_headers = api_headers
   )
-  Chat$new(provider = provider, system_prompt = system_prompt, echo = echo)
+  model_obj <- Model(
+    name = model,
+    params = params,
+    extra_args = api_args
+  )
+  Chat$new(
+    provider = provider,
+    model = model_obj,
+    system_prompt = system_prompt,
+    echo = echo
+  )
 }
 
 chat_azure_openai_test <- function(
diff --git a/R/provider-cloudflare.R b/R/provider-cloudflare.R
index 1c23079c8..0a337977b 100644
--- a/R/provider-cloudflare.R
+++ b/R/provider-cloudflare.R
@@ -58,14 +58,21 @@ chat_cloudflare <- function(
   provider <- ProviderCloudflare(
     name = "Cloudflare",
     base_url = base_url,
-    model = model,
-    params = params,
     credentials = credentials,
-    extra_args = api_args,
     extra_headers = api_headers
   )
+  model_obj <- Model(
+    name = model,
+    params = params,
+    extra_args = api_args
+  )
 
-  Chat$new(provider = provider, system_prompt = system_prompt, echo = echo)
+  Chat$new(
+    provider = provider,
+    model = model_obj,
+    system_prompt = system_prompt,
+    echo = echo
+  )
 }
 
 ProviderCloudflare <- new_class("ProviderCloudflare", parent = ProviderOpenAI)
diff --git a/R/provider-databricks.R b/R/provider-databricks.R
index 9737dbf48..755b113be 100644
--- a/R/provider-databricks.R
+++ b/R/provider-databricks.R
@@ -68,13 +68,20 @@ chat_databricks <- function(
   provider <- ProviderDatabricks(
     name = "Databricks",
     base_url = workspace,
-    model = model,
-    params = params,
-    extra_args = api_args,
     credentials = credentials,
     extra_headers = api_headers
   )
-  Chat$new(provider = provider, system_prompt = system_prompt, echo = echo)
+  model_obj <- Model(
+    name = model,
+    params = params,
+    extra_args = api_args
+  )
+  Chat$new(
+    provider = provider,
+    model = model_obj,
+    system_prompt = system_prompt,
+    echo = echo
+  )
 }
 
 ProviderDatabricks <- new_class(
@@ -93,6 +100,7 @@ method(base_request, ProviderDatabricks) <- function(provider) {
 
 method(chat_body, ProviderDatabricks) <- function(
   provider,
+  model,
   stream = TRUE,
   turns = list(),
   tools = list(),
@@ -100,13 +108,14 @@ method(chat_body, ProviderDatabricks) <- function(
 ) {
   body <- chat_body(
     super(provider, ProviderOpenAI),
+    model,
     stream = stream,
     turns = turns,
     tools = tools,
     type = type
   )
 
-  params <- chat_params(provider, provider@params)
+  params <- chat_params(provider, model)
   body <- modify_list(body, params)
 
   # Databricks doesn't support stream options
@@ -115,10 +124,10 @@ method(chat_body, ProviderDatabricks) <- function(
   body
 }
 
-method(chat_params, ProviderDatabricks) <- function(provider, params) {
+method(chat_params, ProviderDatabricks) <- function(provider, model) {
   # https://docs.databricks.com/aws/en/machine-learning/foundation-model-apis/api-reference#chat-request
   standardise_params(
-    params,
+    model@params,
     c(
       temperature = "temperature",
       top_p = "topP",
diff --git a/R/provider-deepseek.R b/R/provider-deepseek.R
index e66b37e42..5d84acd61 100644
--- a/R/provider-deepseek.R
+++ b/R/provider-deepseek.R
@@ -50,21 +50,28 @@ chat_deepseek <- function(
   provider <- ProviderDeepSeek(
     name = "DeepSeek",
     base_url = base_url,
-    model = model,
-    params = params,
-    extra_args = api_args,
     credentials = credentials,
     extra_headers = api_headers
   )
-  Chat$new(provider = provider, system_prompt = system_prompt, echo = echo)
+  model_obj <- Model(
+    name = model,
+    params = params,
+    extra_args = api_args
+  )
+  Chat$new(
+    provider = provider,
+    model = model_obj,
+    system_prompt = system_prompt,
+    echo = echo
+  )
 }
 
 ProviderDeepSeek <- new_class("ProviderDeepSeek", parent = ProviderOpenAI)
 
-method(chat_params, ProviderDeepSeek) <- function(provider, params) {
+method(chat_params, ProviderDeepSeek) <- function(provider, model) {
   # https://platform.deepseek.com/api-docs/api/create-chat-completion
   standardise_params(
-    params,
+    model@params,
     c(
       frequency_penalty = "frequency_penalty",
       max_tokens = "max_tokens",
diff --git a/R/provider-github.R b/R/provider-github.R
index 379c2170e..e31c9e99e 100644
--- a/R/provider-github.R
+++ b/R/provider-github.R
@@ -91,7 +91,7 @@ models_github <- function(
 
   provider <- ProviderOpenAI(
     name = "github",
-    model = "",
+    base_url = base_url,
     credentials = credentials
   )
 
diff --git a/R/provider-google.R b/R/provider-google.R
index 1bcb9772e..42e38c589 100644
--- a/R/provider-google.R
+++ b/R/provider-google.R
@@ -60,13 +60,21 @@ chat_google_gemini <- function(
   provider <- ProviderGoogleGemini(
     name = "Google/Gemini",
     base_url = base_url,
-    model = model,
-    params = params %||% params(),
-    extra_args = api_args,
     extra_headers = api_headers,
-    credentials = credentials
+    credentials = credentials,
+    model = model
+  )
+  model_obj <- Model(
+    name = model,
+    params = params %||% params(),
+    extra_args = api_args
+  )
+  Chat$new(
+    provider = provider,
+    model = model_obj,
+    system_prompt = system_prompt,
+    echo = echo
   )
-  Chat$new(provider = provider, system_prompt = system_prompt, echo = echo)
 }
 
 chat_google_gemini_test <- function(
@@ -107,13 +115,21 @@ chat_google_vertex <- function(
   provider <- ProviderGoogleGemini(
     name = "Google/Vertex",
     base_url = vertex_url(location, project_id),
-    model = model,
-    params = params %||% params(),
-    extra_args = api_args,
     extra_headers = api_headers,
-    credentials = credentials
+    credentials = credentials,
+    model = model
+  )
+  model_obj <- Model(
+    name = model,
+    params = params %||% params(),
+    extra_args = api_args
+  )
+  Chat$new(
+    provider = provider,
+    model = model_obj,
+    system_prompt = system_prompt,
+    echo = echo
   )
-  Chat$new(provider = provider, system_prompt = system_prompt, echo = echo)
 }
 
 # https://cloud.google.com/vertex-ai/docs/reference/rest/v1/projects.locations.endpoints/generateContent
@@ -153,6 +169,7 @@ method(base_request, ProviderGoogleGemini) <- function(provider) {
 
 method(chat_request, ProviderGoogleGemini) <- function(
   provider,
+  model,
   stream = TRUE,
   turns = list(),
   tools = list(),
@@ -179,12 +196,13 @@ method(chat_request, ProviderGoogleGemini) <- function(
 
   body <- chat_body(
     provider = provider,
+    model = model,
     stream = stream,
     turns = turns,
     tools = tools,
     type = type
   )
-  body <- modify_list(body, provider@extra_args)
+  body <- modify_list(body, model@extra_args)
 
   req <- req_body_json(req, body)
   req <- req_headers(req, !!!provider@extra_headers)
@@ -193,6 +211,7 @@ method(chat_request, ProviderGoogleGemini) <- function(
 
 method(chat_body, ProviderGoogleGemini) <- function(
   provider,
+  model,
   stream = TRUE,
   turns = list(),
   tools = list(),
@@ -204,7 +223,7 @@ method(chat_body, ProviderGoogleGemini) <- function(
     system <- list(parts = list(text = ""))
   }
 
-  generation_config <- chat_params(provider, provider@params)
+  generation_config <- chat_params(provider, model)
   if (!is.null(type)) {
     generation_config$response_mime_type <- "application/json"
     generation_config$response_schema <- as_json(provider, type)
@@ -236,9 +255,9 @@ method(chat_body, ProviderGoogleGemini) <- function(
   ))
 }
 
-method(chat_params, ProviderGoogleGemini) <- function(provider, params) {
+method(chat_params, ProviderGoogleGemini) <- function(provider, model) {
   standardise_params(
-    params,
+    model@params,
     c(
       temperature = "temperature",
       topP = "top_p",
@@ -278,7 +297,7 @@ method(stream_merge_chunks, ProviderGoogleGemini) <- function(
   }
 }
 
-method(value_tokens, ProviderGoogleGemini) <- function(provider, json) {
+method(value_tokens, ProviderGoogleGemini) <- function(provider, model, json) {
   usage <- json$usageMetadata
   cached <- usage$cachedContentTokenCount %||% 0
 
@@ -291,6 +310,7 @@ method(value_tokens, ProviderGoogleGemini) <- function(provider, json) {
 
 method(value_turn, ProviderGoogleGemini) <- function(
   provider,
+  model,
   result,
   has_type = FALSE
 ) {
@@ -322,8 +342,8 @@ method(value_turn, ProviderGoogleGemini) <- function(
     }
   })
   contents <- compact(contents)
-  tokens <- value_tokens(provider, result)
-  cost <- get_token_cost(provider, tokens)
+  tokens <- value_tokens(provider, model, result)
+  cost <- get_token_cost(provider, model, tokens)
   AssistantTurn(contents, json = result, tokens = unlist(tokens), cost = cost)
 }
 
diff --git a/R/provider-groq.R b/R/provider-groq.R
index 48e3c9ae6..48a5224b5 100644
--- a/R/provider-groq.R
+++ b/R/provider-groq.R
@@ -53,13 +53,20 @@ chat_groq <- function(
   provider <- ProviderGroq(
     name = "Groq",
     base_url = base_url,
-    model = model,
-    params = params,
-    extra_args = api_args,
     credentials = credentials,
     extra_headers = api_headers
   )
-  Chat$new(provider = provider, system_prompt = system_prompt, echo = echo)
+  model_obj <- Model(
+    name = model,
+    params = params,
+    extra_args = api_args
+  )
+  Chat$new(
+    provider = provider,
+    model = model_obj,
+    system_prompt = system_prompt,
+    echo = echo
+  )
 }
 
 ProviderGroq <- new_class("ProviderGroq", parent = ProviderOpenAI)
diff --git a/R/provider-huggingface.R b/R/provider-huggingface.R
index f291983cd..554002bf1 100644
--- a/R/provider-huggingface.R
+++ b/R/provider-huggingface.R
@@ -56,13 +56,20 @@ chat_huggingface <- function(
   provider <- ProviderHuggingFace(
     name = "HuggingFace",
     base_url = base_url,
-    model = model,
-    params = params,
-    extra_args = api_args,
     credentials = credentials,
     extra_headers = api_headers
   )
-  Chat$new(provider = provider, system_prompt = system_prompt, echo = echo)
+  model_obj <- Model(
+    name = model,
+    params = params,
+    extra_args = api_args
+  )
+  Chat$new(
+    provider = provider,
+    model = model_obj,
+    system_prompt = system_prompt,
+    echo = echo
+  )
 }
 
 ProviderHuggingFace <- new_class("ProviderHuggingFace", parent = ProviderOpenAI)
diff --git a/R/provider-mistral.R b/R/provider-mistral.R
index a063e9f7a..138588f96 100644
--- a/R/provider-mistral.R
+++ b/R/provider-mistral.R
@@ -44,13 +44,20 @@ chat_mistral <- function(
   provider <- ProviderMistral(
     name = "Mistral",
     base_url = mistral_base_url,
-    model = model,
-    params = params,
-    extra_args = api_args,
     credentials = credentials,
     extra_headers = api_headers
   )
-  Chat$new(provider = provider, system_prompt = system_prompt, echo = echo)
+  model_obj <- Model(
+    name = model,
+    params = params,
+    extra_args = api_args
+  )
+  Chat$new(
+    provider = provider,
+    model = model_obj,
+    system_prompt = system_prompt,
+    echo = echo
+  )
 }
 
 mistral_base_url <- "https://api.mistral.ai/v1/"
@@ -91,6 +98,7 @@ method(base_request, ProviderMistral) <- function(provider) {
 
 method(chat_body, ProviderMistral) <- function(
   provider,
+  model,
   stream = TRUE,
   turns = list(),
   tools = list(),
@@ -98,6 +106,7 @@ method(chat_body, ProviderMistral) <- function(
 ) {
   body <- chat_body(
     super(provider, ProviderOpenAI),
+    model,
     stream = stream,
     turns = turns,
     tools = tools,
@@ -110,9 +119,9 @@ method(chat_body, ProviderMistral) <- function(
   body
 }
 
-method(chat_params, ProviderMistral) <- function(provider, params) {
+method(chat_params, ProviderMistral) <- function(provider, model) {
   standardise_params(
-    params,
+    model@params,
     c(
       temperature = "temperature",
       top_p = "top_p",
@@ -137,7 +146,6 @@ mistral_key <- function() {
 models_mistral <- function(api_key = mistral_key()) {
   provider <- ProviderMistral(
     name = "Mistral",
-    model = "",
     base_url = mistral_base_url,
     credentials = function() api_key
   )
diff --git a/R/provider-ollama.R b/R/provider-ollama.R
index eb10b475a..9f7b0c9f6 100644
--- a/R/provider-ollama.R
+++ b/R/provider-ollama.R
@@ -85,14 +85,22 @@ chat_ollama <- function(
   provider <- ProviderOllama(
     name = "Ollama",
     base_url = file.path(base_url, "v1"), ## the v1 portion of the path is added for openAI compatible API
-    model = model,
-    params = params %||% params(),
-    extra_args = api_args,
     credentials = credentials,
-    extra_headers = api_headers
+    extra_headers = api_headers,
+    model = model
+  )
+  model_obj <- Model(
+    name = model,
+    params = params %||% params(),
+    extra_args = api_args
   )
 
-  Chat$new(provider = provider, system_prompt = system_prompt, echo = echo)
+  Chat$new(
+    provider = provider,
+    model = model_obj,
+    system_prompt = system_prompt,
+    echo = echo
+  )
 }
 
 ProviderOllama <- new_class(
@@ -103,10 +111,10 @@ ProviderOllama <- new_class(
   )
 )
 
-method(chat_params, ProviderOllama) <- function(provider, params) {
+method(chat_params, ProviderOllama) <- function(provider, model) {
   # https://github.com/ollama/ollama/blob/main/docs/api.md#generate-a-chat-completion
   standardise_params(
-    params,
+    model@params,
     c(
       frequency_penalty = "frequency_penalty",
       presence_penalty = "presence_penalty",
diff --git a/R/provider-openai-responses.R b/R/provider-openai-responses.R
index d082e7ec3..c06ef4f10 100644
--- a/R/provider-openai-responses.R
+++ b/R/provider-openai-responses.R
@@ -55,14 +55,21 @@ chat_openai_responses <- function(
   provider <- ProviderOpenAIResponses(
     name = "OpenAI",
     base_url = base_url,
-    model = model,
-    params = params %||% params(),
-    extra_args = api_args,
     extra_headers = api_headers,
     credentials = credentials,
     service_tier = service_tier
   )
-  Chat$new(provider = provider, system_prompt = system_prompt, echo = echo)
+  model_obj <- Model(
+    name = model,
+    params = params %||% params(),
+    extra_args = api_args
+  )
+  Chat$new(
+    provider = provider,
+    model = model_obj,
+    system_prompt = system_prompt,
+    echo = echo
+  )
 }
 chat_openai_responses_test <- function(
   system_prompt = "Be terse.",
@@ -101,6 +108,7 @@ method(chat_path, ProviderOpenAIResponses) <- function(provider) {
 # https://platform.openai.com/docs/api-reference/responses
 method(chat_body, ProviderOpenAIResponses) <- function(
   provider,
+  model,
   stream = TRUE,
   turns = list(),
   tools = list(),
@@ -124,7 +132,7 @@ method(chat_body, ProviderOpenAIResponses) <- function(
   }
 
   # https://platform.openai.com/docs/api-reference/responses/create#responses-create-include
-  params <- chat_params(provider, provider@params)
+  params <- chat_params(provider, model)
 
   if (has_name(params, "reasoning_effort")) {
     reasoning <- list(
@@ -157,9 +165,9 @@ method(chat_body, ProviderOpenAIResponses) <- function(
 }
 
 
-method(chat_params, ProviderOpenAIResponses) <- function(provider, params) {
+method(chat_params, ProviderOpenAIResponses) <- function(provider, model) {
   standardise_params(
-    params,
+    model@params,
     c(
       temperature = "temperature",
       top_p = "top_p",
@@ -205,7 +213,11 @@ method(stream_merge_chunks, ProviderOpenAIResponses) <- function(
   }
 }
 
-method(value_tokens, ProviderOpenAIResponses) <- function(provider, json) {
+method(value_tokens, ProviderOpenAIResponses) <- function(
+  provider,
+  model,
+  json
+) {
   usage <- json$usage
   cached_tokens <- usage$input_tokens_details$cached_tokens %||% 0
 
diff --git a/R/provider-openai.R b/R/provider-openai.R
index 959799298..8e5f5b0d3 100644
--- a/R/provider-openai.R
+++ b/R/provider-openai.R
@@ -69,13 +69,20 @@ chat_openai <- function(
   provider <- ProviderOpenAI(
     name = "OpenAI",
     base_url = base_url,
-    model = model,
-    params = params %||% params(),
-    extra_args = api_args,
     extra_headers = api_headers,
     credentials = credentials
   )
-  Chat$new(provider = provider, system_prompt = system_prompt, echo = echo)
+  model <- Model(
+    name = model,
+    params = params %||% params(),
+    extra_args = api_args
+  )
+  Chat$new(
+    provider = provider,
+    model = model,
+    system_prompt = system_prompt,
+    echo = echo
+  )
 }
 chat_openai_test <- function(
   system_prompt = "Be terse.",
@@ -147,6 +154,7 @@ method(chat_path, ProviderOpenAI) <- function(provider) {
 # https://platform.openai.com/docs/api-reference/chat/create
 method(chat_body, ProviderOpenAI) <- function(
   provider,
+  model,
   stream = TRUE,
   turns = list(),
   tools = list(),
@@ -168,11 +176,11 @@ method(chat_body, ProviderOpenAI) <- function(
     response_format <- NULL
   }
 
-  params <- chat_params(provider, provider@params)
+  params <- chat_params(provider, model)
 
   compact(list2(
     messages = messages,
-    model = provider@model,
+    model = model@name,
     !!!params,
     stream = stream,
     stream_options = if (stream) list(include_usage = TRUE),
@@ -182,9 +190,9 @@ method(chat_body, ProviderOpenAI) <- function(
 }
 
 
-method(chat_params, ProviderOpenAI) <- function(provider, params) {
+method(chat_params, ProviderOpenAI) <- function(provider, model) {
   standardise_params(
-    params,
+    model@params,
     c(
       temperature = "temperature",
       top_p = "top_p",
@@ -227,7 +235,7 @@ method(stream_merge_chunks, ProviderOpenAI) <- function(
   }
 }
 
-method(value_tokens, ProviderOpenAI) <- function(provider, json) {
+method(value_tokens, ProviderOpenAI) <- function(provider, model, json) {
   usage <- json$usage
   cached_tokens <- usage$prompt_tokens_details$cached_tokens %||% 0
 
@@ -240,6 +248,7 @@ method(value_tokens, ProviderOpenAI) <- function(provider, json) {
 
 method(value_turn, ProviderOpenAI) <- function(
   provider,
+  model,
   result,
   has_type = FALSE
 ) {
@@ -272,8 +281,8 @@ method(value_turn, ProviderOpenAI) <- function(
     content <- c(content, calls)
   }
 
-  tokens <- value_tokens(provider, result)
-  cost <- get_token_cost(provider, tokens)
+  tokens <- value_tokens(provider, model, result)
+  cost <- get_token_cost(provider, model, tokens)
   AssistantTurn(content, json = result, tokens = unlist(tokens), cost = cost)
 }
 
@@ -525,11 +534,12 @@ method(batch_retrieve, ProviderOpenAI) <- function(provider, batch) {
 
 method(batch_result_turn, ProviderOpenAI) <- function(
   provider,
+  model,
   result,
   has_type = FALSE
 ) {
   if (result$status_code == 200) {
-    value_turn(provider, result$body, has_type = has_type)
+    value_turn(provider, model, result$body, has_type = has_type)
   } else {
     NULL
   }
@@ -553,7 +563,6 @@ models_openai <- function(
 
   provider <- ProviderOpenAI(
     name = "OpenAI",
-    model = "",
     base_url = base_url,
     credentials = credentials
   )
diff --git a/R/provider-openrouter.R b/R/provider-openrouter.R
index 928f5bfe9..62a6fd5a4 100644
--- a/R/provider-openrouter.R
+++ b/R/provider-openrouter.R
@@ -47,13 +47,20 @@ chat_openrouter <- function(
   provider <- ProviderOpenRouter(
     name = "OpenRouter",
     base_url = "https://openrouter.ai/api/v1",
-    model = model,
-    params = params,
-    extra_args = api_args,
     credentials = credentials,
     extra_headers = api_headers
   )
-  Chat$new(provider = provider, system_prompt = system_prompt, echo = echo)
+  model_obj <- Model(
+    name = model,
+    params = params,
+    extra_args = api_args
+  )
+  Chat$new(
+    provider = provider,
+    model = model_obj,
+    system_prompt = system_prompt,
+    echo = echo
+  )
 }
 
 chat_openrouter_test <- function(..., echo = "none") {
@@ -65,10 +72,10 @@ ProviderOpenRouter <- new_class(
   parent = ProviderOpenAI,
 )
 
-method(chat_params, ProviderOpenRouter) <- function(provider, params) {
+method(chat_params, ProviderOpenRouter) <- function(provider, model) {
   # https://openrouter.ai/docs/api-reference/parameters
   standardise_params(
-    params,
+    model@params,
     c(
       temperature = "temperature",
       top_p = "top_p",
diff --git a/R/provider-perplexity.R b/R/provider-perplexity.R
index 73274a452..f76e36b1c 100644
--- a/R/provider-perplexity.R
+++ b/R/provider-perplexity.R
@@ -53,13 +53,20 @@ chat_perplexity <- function(
   provider <- ProviderPerplexity(
     name = "Perplexity",
     base_url = base_url,
-    model = model,
-    params = params,
-    extra_args = api_args,
     credentials = credentials,
     extra_headers = api_headers
   )
-  Chat$new(provider = provider, system_prompt = system_prompt, echo = echo)
+  model_obj <- Model(
+    name = model,
+    params = params,
+    extra_args = api_args
+  )
+  Chat$new(
+    provider = provider,
+    model = model_obj,
+    system_prompt = system_prompt,
+    echo = echo
+  )
 }
 
 ProviderPerplexity <- new_class(
@@ -67,10 +74,10 @@ ProviderPerplexity <- new_class(
   parent = ProviderOpenAI,
 )
 
-method(chat_params, ProviderPerplexity) <- function(provider, params) {
+method(chat_params, ProviderPerplexity) <- function(provider, model) {
   # https://docs.perplexity.ai/api-reference/chat-completions-post
   standardise_params(
-    params,
+    model@params,
     c(
       max_tokens = "max_tokens",
       temperature = "temperature",
diff --git a/R/provider-portkey.R b/R/provider-portkey.R
index e00c73cf5..40f5d3381 100644
--- a/R/provider-portkey.R
+++ b/R/provider-portkey.R
@@ -50,14 +50,21 @@ chat_portkey <- function(
   provider <- ProviderPortkeyAI(
     name = "PortkeyAI",
     base_url = base_url,
-    model = model,
-    params = params,
-    extra_args = api_args,
     credentials = credentials,
     virtual_key = virtual_key,
     extra_headers = api_headers
   )
-  Chat$new(provider = provider, system_prompt = system_prompt, echo = echo)
+  model_obj <- Model(
+    name = model,
+    params = params,
+    extra_args = api_args
+  )
+  Chat$new(
+    provider = provider,
+    model = model_obj,
+    system_prompt = system_prompt,
+    echo = echo
+  )
 }
 
 chat_portkey_test <- function(
@@ -122,7 +129,6 @@ models_portkey <- function(
 ) {
   provider <- ProviderPortkeyAI(
     name = "PortkeyAI",
-    model = "",
     base_url = base_url,
     credentials = function() api_key,
     virtual_key = virtual_key
diff --git a/R/provider-snowflake.R b/R/provider-snowflake.R
index 24c2dd010..562e9e2ba 100644
--- a/R/provider-snowflake.R
+++ b/R/provider-snowflake.R
@@ -61,13 +61,20 @@ chat_snowflake <- function(
     base_url = snowflake_url(account),
     account = account,
     credentials = credentials,
-    model = model,
-    params = params,
-    extra_args = api_args,
     extra_headers = api_headers
   )
+  model_obj <- Model(
+    name = model,
+    params = params,
+    extra_args = api_args
+  )
 
-  Chat$new(provider = provider, system_prompt = system_prompt, echo = echo)
+  Chat$new(
+    provider = provider,
+    model = model_obj,
+    system_prompt = system_prompt,
+    echo = echo
+  )
 }
 
 ProviderSnowflakeCortex <- new_class(
@@ -100,6 +107,7 @@ method(chat_path, ProviderSnowflakeCortex) <- function(provider) {
 # See: https://docs.snowflake.com/en/user-guide/snowflake-cortex/cortex-llm-rest-api#api-reference
 method(chat_body, ProviderSnowflakeCortex) <- function(
   provider,
+  model,
   stream = TRUE,
   turns = list(),
   tools = list(),
@@ -115,10 +123,10 @@ method(chat_body, ProviderSnowflakeCortex) <- function(
     response_format <- NULL
   }
 
-  params <- chat_params(provider, provider@params)
+  params <- chat_params(provider, model)
   compact(list2(
     messages = messages,
-    model = provider@model,
+    model = model@name,
     !!!params,
     stream = stream,
     tools = tools,
@@ -145,9 +153,9 @@ method(as_json, list(ProviderSnowflakeCortex, TypeObject)) <- function(
 }
 
 # See: https://docs.snowflake.com/en/user-guide/snowflake-cortex/cortex-llm-rest-api#optional-json-arguments
-method(chat_params, ProviderSnowflakeCortex) <- function(provider, params) {
+method(chat_params, ProviderSnowflakeCortex) <- function(provider, model) {
   standardise_params(
-    params,
+    model@params,
     c(
       temperature = "temperature",
       top_p = "top_p",
@@ -215,7 +223,11 @@ method(stream_merge_chunks, ProviderSnowflakeCortex) <- function(
   result
 }
 
-method(value_tokens, ProviderSnowflakeCortex) <- function(provider, json) {
+method(value_tokens, ProviderSnowflakeCortex) <- function(
+  provider,
+  model,
+  json
+) {
   tokens(
     input = json$usage$prompt_tokens,
     output = json$usage$completion_tokens
@@ -224,6 +236,7 @@ method(value_tokens, ProviderSnowflakeCortex) <- function(provider, json) {
 
 method(value_turn, ProviderSnowflakeCortex) <- function(
   provider,
+  model,
   result,
   has_type = FALSE
 ) {
@@ -252,8 +265,8 @@ method(value_turn, ProviderSnowflakeCortex) <- function(
       )
     }
   })
-  tokens <- value_tokens(provider, result)
-  cost <- get_token_cost(provider, tokens)
+  tokens <- value_tokens(provider, model, result)
+  cost <- get_token_cost(provider, model, tokens)
   AssistantTurn(contents, json = result, tokens = unlist(tokens), cost = cost)
 }
 
diff --git a/R/provider-vllm.R b/R/provider-vllm.R
index 39da5d744..254f0a3b9 100644
--- a/R/provider-vllm.R
+++ b/R/provider-vllm.R
@@ -57,13 +57,20 @@ chat_vllm <- function(
   provider <- ProviderVllm(
     name = "VLLM",
     base_url = base_url,
-    model = model,
-    params = params,
-    extra_args = api_args,
     credentials = credentials,
     extra_headers = api_headers
   )
-  Chat$new(provider = provider, system_prompt = system_prompt, echo = echo)
+  model_obj <- Model(
+    name = model,
+    params = params,
+    extra_args = api_args
+  )
+  Chat$new(
+    provider = provider,
+    model = model_obj,
+    system_prompt = system_prompt,
+    echo = echo
+  )
 }
 
 chat_vllm_test <- function(..., echo = "none") {
diff --git a/R/provider.R b/R/provider.R
index 360433b5d..9f7744737 100644
--- a/R/provider.R
+++ b/R/provider.R
@@ -14,36 +14,29 @@ NULL
 #'
 #' @export
 #' @param name Name of the provider.
-#' @param model Name of the model.
 #' @param base_url The base URL for the API.
-#' @param params A list of standard parameters created by [params()].
 #' @param credentials A zero-argument function that returns the credentials to use
 #'   for authentication. Can either return a string, representing an API key,
 #'   or a named list of headers.
-#' @param extra_args Arbitrary extra arguments to be included in the request body.
 #' @param extra_headers Arbitrary extra headers to be added to the request.
 #' @return An S7 Provider object.
 #' @examples
 #' Provider(
 #'   name = "CoolModels",
-#'   model = "my_model",
 #'   base_url = "https://cool-models.com"
 #' )
 Provider <- new_class(
   "Provider",
   properties = list(
     name = prop_string(),
-    model = prop_string(),
     base_url = prop_string(),
-    params = class_list,
-    extra_args = class_list,
     extra_headers = class_character,
     credentials = class_function | NULL
   )
 )
 
-test_provider <- function(name = "", model = "", base_url = "", ...) {
-  Provider(name = name, model = model, base_url = base_url, ...)
+test_provider <- function(name = "", base_url = "", ...) {
+  Provider(name = name, base_url = base_url, ...)
 }
 
 # Create a request------------------------------------
@@ -65,6 +58,7 @@ chat_request <- new_generic(
   "provider",
   function(
     provider,
+    model,
     stream = TRUE,
     turns = list(),
     tools = list(),
@@ -76,6 +70,7 @@ chat_request <- new_generic(
 
 method(chat_request, Provider) <- function(
   provider,
+  model,
   stream = TRUE,
   turns = list(),
   tools = list(),
@@ -86,12 +81,13 @@ method(chat_request, Provider) <- function(
 
   body <- chat_body(
     provider = provider,
+    model = model,
     stream = stream,
     turns = turns,
     tools = tools,
     type = type
   )
-  body <- modify_list(body, provider@extra_args)
+  body <- modify_list(body, model@extra_args)
   req <- req_body_json(req, body)
   req <- req_headers(req, !!!provider@extra_headers)
 
@@ -103,6 +99,7 @@ chat_body <- new_generic(
   "provider",
   function(
     provider,
+    model,
     stream = TRUE,
     turns = list(),
     tools = list(),
@@ -130,7 +127,7 @@ method(chat_resp_stream, Provider) <- function(provider, resp) {
 chat_params <- new_generic(
   "chat_params",
   "provider",
-  function(provider, params) {
+  function(provider, model) {
     S7_dispatch()
   }
 )
@@ -161,18 +158,24 @@ stream_merge_chunks <- new_generic(
 
 # Extract data from non-streaming results --------------------------------------
 
-value_turn <- new_generic("value_turn", "provider")
+value_turn <- new_generic(
+  "value_turn",
+  "provider",
+  function(provider, model, json, ...) {
+    S7_dispatch()
+  }
+)
 
 # Extract token counts from API response
 # Returns a named list produced by token_usage()
 value_tokens <- new_generic(
   "value_tokens",
   "provider",
-  function(provider, json) {
+  function(provider, model, json) {
     S7_dispatch()
   }
 )
-method(value_tokens, Provider) <- function(provider, json) {
+method(value_tokens, Provider) <- function(provider, model, json) {
   tokens()
 }
 
@@ -258,7 +261,7 @@ batch_retrieve <- new_generic(
 batch_result_turn <- new_generic(
   "batch_result_turn",
   "provider",
-  function(provider, result, has_type = FALSE) {
+  function(provider, model, result, has_type = FALSE) {
     S7_dispatch()
   }
 )
diff --git a/R/tokens.R b/R/tokens.R
index c43c00432..0f4e59133 100644
--- a/R/tokens.R
+++ b/R/tokens.R
@@ -20,11 +20,11 @@ map_tokens <- function(x, f, ...) {
   out
 }
 
-log_tokens <- function(provider, tokens, cost) {
+log_tokens <- function(provider, model, tokens, cost) {
   i <- vctrs::vec_match(
     data.frame(
       provider = provider@name,
-      model = provider@model
+      model = model@name
     ),
     the$tokens[c("provider", "model")]
   )
@@ -32,7 +32,7 @@ log_tokens <- function(provider, tokens, cost) {
   if (is.na(i)) {
     new_row <- tokens_row(
       provider@name,
-      provider@model,
+      model@name,
       tokens$input,
       tokens$output,
       tokens$cached_input,
@@ -50,14 +50,14 @@ log_tokens <- function(provider, tokens, cost) {
   invisible()
 }
 
-log_turn <- function(provider, turn) {
-  log_tokens(provider, exec(tokens, !!!as.list(turn@tokens)), turn@cost)
+log_turn <- function(provider, model, turn) {
+  log_tokens(provider, model, exec(tokens, !!!as.list(turn@tokens)), turn@cost)
 }
 
-log_turns <- function(provider, turns) {
+log_turns <- function(provider, model, turns) {
   for (turn in turns) {
     if (S7_inherits(turn, AssistantTurn)) {
-      log_turn(provider, turn)
+      log_turn(provider, model, turn)
     }
   }
 }
@@ -110,14 +110,14 @@ token_usage <- function() {
 # Cost ----------------------------------------------------------------------
 
 has_cost <- function(provider, model) {
-  needle <- data.frame(provider = provider@name, model = model)
+  needle <- data.frame(provider = provider@name, model = model@name)
   vctrs::vec_in(needle, prices[c("provider", "model")])
 }
 
-get_token_cost <- function(provider, tokens, variant = "") {
+get_token_cost <- function(provider, model, tokens, variant = "") {
   needle <- data.frame(
     provider = provider@name,
-    model = provider@model,
+    model = model@name,
     variant = variant
   )
   idx <- vctrs::vec_match(needle, prices[c("provider", "model", "variant")])
diff --git a/man/Chat.Rd b/man/Chat.Rd
index 66ec02d58..7f1d6bb1b 100644
--- a/man/Chat.Rd
+++ b/man/Chat.Rd
@@ -45,6 +45,7 @@ chat$chat("Tell me a funny joke")
 \item \href{#method-Chat-register_tool}{\code{Chat$register_tool()}}
 \item \href{#method-Chat-register_tools}{\code{Chat$register_tools()}}
 \item \href{#method-Chat-get_provider}{\code{Chat$get_provider()}}
+\item \href{#method-Chat-get_model_obj}{\code{Chat$get_model_obj()}}
 \item \href{#method-Chat-get_tools}{\code{Chat$get_tools()}}
 \item \href{#method-Chat-set_tools}{\code{Chat$set_tools()}}
 \item \href{#method-Chat-on_tool_request}{\code{Chat$on_tool_request()}}
@@ -57,7 +58,7 @@ chat$chat("Tell me a funny joke")
 \if{latex}{\out{\hypertarget{method-Chat-new}{}}}
 \subsection{Method \code{new()}}{
 \subsection{Usage}{
-\if{html}{\out{<div class="r">}}\preformatted{Chat$new(provider, system_prompt = NULL, echo = "none")}\if{html}{\out{</div>}}
+\if{html}{\out{<div class="r">}}\preformatted{Chat$new(provider, model, system_prompt = NULL, echo = "none")}\if{html}{\out{</div>}}
 }
 
 \subsection{Arguments}{
@@ -65,6 +66,8 @@ chat$chat("Tell me a funny joke")
 \describe{
 \item{\code{provider}}{A provider object.}
 
+\item{\code{model}}{A model object.}
+
 \item{\code{system_prompt}}{System prompt to start the conversation with.}
 
 \item{\code{echo}}{One of the following options:
@@ -447,6 +450,16 @@ Get the underlying provider object. For expert use only.
 \if{html}{\out{<div class="r">}}\preformatted{Chat$get_provider()}\if{html}{\out{</div>}}
 }
 
+}
+\if{html}{\out{<hr>}}
+\if{html}{\out{<a id="method-Chat-get_model_obj"></a>}}
+\if{latex}{\out{\hypertarget{method-Chat-get_model_obj}{}}}
+\subsection{Method \code{get_model_obj()}}{
+Get the underlying model object. For expert use only.
+\subsection{Usage}{
+\if{html}{\out{<div class="r">}}\preformatted{Chat$get_model_obj()}\if{html}{\out{</div>}}
+}
+
 }
 \if{html}{\out{<hr>}}
 \if{html}{\out{<a id="method-Chat-get_tools"></a>}}
diff --git a/man/Model.Rd b/man/Model.Rd
new file mode 100644
index 000000000..63d9a1a02
--- /dev/null
+++ b/man/Model.Rd
@@ -0,0 +1,35 @@
+% Generated by roxygen2: do not edit by hand
+% Please edit documentation in R/model.R
+\name{Model}
+\alias{Model}
+\title{A chatbot model}
+\usage{
+Model(name = stop("Required"), params = list(), extra_args = list())
+}
+\arguments{
+\item{name}{Name of the model (e.g., "gpt-4", "claude-sonnet-4").}
+
+\item{params}{A list of standard parameters created by \code{\link[=params]{params()}}.}
+
+\item{extra_args}{Arbitrary extra arguments to be included in the request body.}
+}
+\value{
+An S7 Model object.
+}
+\description{
+A Model captures the details of a specific language model and its
+configuration parameters. This includes the model name, parameters like
+temperature and max_tokens, and any extra arguments to include in API
+requests.
+}
+\details{
+Model objects are typically created internally by \code{\link[=chat_openai]{chat_openai()}},
+\code{\link[=chat_anthropic]{chat_anthropic()}}, and other \verb{chat_*()} functions. You generally won't
+need to create them directly unless you're implementing a custom provider.
+}
+\examples{
+Model(
+  name = "gpt-4",
+  params = params(temperature = 0.7)
+)
+}
diff --git a/man/Provider.Rd b/man/Provider.Rd
index 38187849a..1a6e9c14b 100644
--- a/man/Provider.Rd
+++ b/man/Provider.Rd
@@ -6,10 +6,7 @@
 \usage{
 Provider(
   name = stop("Required"),
-  model = stop("Required"),
   base_url = stop("Required"),
-  params = list(),
-  extra_args = list(),
   extra_headers = character(0),
   credentials = function() NULL
 )
@@ -17,14 +14,8 @@ Provider(
 \arguments{
 \item{name}{Name of the provider.}
 
-\item{model}{Name of the model.}
-
 \item{base_url}{The base URL for the API.}
 
-\item{params}{A list of standard parameters created by \code{\link[=params]{params()}}.}
-
-\item{extra_args}{Arbitrary extra arguments to be included in the request body.}
-
 \item{extra_headers}{Arbitrary extra headers to be added to the request.}
 
 \item{credentials}{A zero-argument function that returns the credentials to use
@@ -48,7 +39,6 @@ the various generics that control the behavior of each provider.
 \examples{
 Provider(
   name = "CoolModels",
-  model = "my_model",
   base_url = "https://cool-models.com"
 )
 }
diff --git a/tests/testthat/testthat-problems.rds b/tests/testthat/testthat-problems.rds
new file mode 100644
index 000000000..af2df6f3e
Binary files /dev/null and b/tests/testthat/testthat-problems.rds differ