Revert

docs/src/manual/models.md

@@ -183,11 +183,10 @@ julia> model = Model(solver);
 By default, the coefficients of affine and quadratic expressions are numbers
 of type either `Float64` or `Complex{Float64}` (see [Complex number support](@ref)).
 
-The type `Float64` can be changed using the `value_type` argument of
-[`Model`](@ref):
+The type `Float64` can be changed using the [`GenericModel`](@ref) constructor:
 
 ```jldoctest
-julia> model = Model(; value_type = Rational{BigInt});
+julia> model = GenericModel{Rational{BigInt}}();
 
 julia> @variable(model, x)
 x

docs/src/tutorials/conic/arbitrary_precision.jl

@@ -16,11 +16,10 @@ import Clarabel
 
 # ## Higher-precision arithmetic
 
-# To create a model with a number type other than `Float64`, use the `value_type`
-# argument to [`Model`](@ref) with an optimizer which supports the same number
-# type:
+# To create a model with a number type other than `Float64`, use [`GenericModel`](@ref)
+# with an optimizer which supports the same number type:
 
-model = Model(Clarabel.Optimizer{BigFloat}; value_type = BigFloat)
+model = GenericModel{BigFloat}(Clarabel.Optimizer{BigFloat})
 
 # The syntax for adding decision variables is the same as a normal JuMP
 # model, except that values are converted to `BigFloat`:
@@ -109,7 +108,7 @@ value.(x) .- [3 // 7, 3 // 14]
 # JuMP also supports solvers with exact rational arithmetic. One example is
 # CDDLib.jl, which supports the `Rational{BigInt}` number type:
 
-model = Model(CDDLib.Optimizer{Rational{BigInt}}; value_type = Rational{BigInt})
+model = GenericModel{Rational{BigInt}}(CDDLib.Optimizer{Rational{BigInt}})
 
 # As before, we can create variables using rational bounds:
 

src/JuMP.jl

@@ -184,10 +184,6 @@ function GenericModel{T}(
  return model
 end
 
-function GenericModel{T}(args...; kwargs...) where {T}
- return Model(args...; value_type = T, kwargs...)
-end
-
 """
  direct_generic_model(
  value_type::Type{T},