diff --git a/.env.example b/.env.example
index f43a82e81..bd9b8bde0 100644
--- a/.env.example
+++ b/.env.example
@@ -11,6 +11,7 @@ GOOGLE_CLOUD_LOCATION=global
 GOOGLE_CLOUD_PROJECT=$(op read "op://RubyLLM/Google Cloud/project")
 GPUSTACK_API_BASE=http://localhost:11444/v1
 GPUSTACK_API_KEY=$(op read "op://RubyLLM/GPUStack/credential")
+MINIMAX_API_KEY=$(op read "op://RubyLLM/MiniMax/credential")
 MISTRAL_API_KEY=$(op read "op://RubyLLM/Mistral/credential")
 OLLAMA_API_BASE=http://localhost:11434/v1
 OPENAI_API_KEY=$(op read "op://RubyLLM/OpenAI/credential")
diff --git a/README.md b/README.md
index 40bd89c95..4746c96cf 100644
--- a/README.md
+++ b/README.md
@@ -140,7 +140,7 @@ response = chat.with_schema(ProductSchema).ask "Analyze this product", with: "pr
 * **Async:** Fiber-based concurrency
 * **Model registry:** 800+ models with capability detection and pricing
 * **Extended thinking:** Control, view, and persist model deliberation
-* **Providers:** OpenAI, xAI, Anthropic, Gemini, VertexAI, Bedrock, DeepSeek, Mistral, Ollama, OpenRouter, Perplexity, GPUStack, and any OpenAI-compatible API
+* **Providers:** OpenAI, xAI, Anthropic, Gemini, VertexAI, Bedrock, DeepSeek, MiniMax, Mistral, Ollama, OpenRouter, Perplexity, GPUStack, and any OpenAI-compatible API
 
 ## Installation
 
diff --git a/docs/_getting_started/configuration.md b/docs/_getting_started/configuration.md
index b5686cb2e..fd0fb5ae6 100644
--- a/docs/_getting_started/configuration.md
+++ b/docs/_getting_started/configuration.md
@@ -76,6 +76,10 @@ RubyLLM.configure do |config|
   config.gpustack_api_base = ENV['GPUSTACK_API_BASE']
   config.gpustack_api_key = ENV['GPUSTACK_API_KEY']
 
+  # MiniMax
+  config.minimax_api_key = ENV['MINIMAX_API_KEY']
+  config.minimax_api_base = ENV['MINIMAX_API_BASE'] # Optional custom MiniMax endpoint (defaults to https://api.minimax.io/v1)
+
   # Mistral
   config.mistral_api_key = ENV['MISTRAL_API_KEY']
 
@@ -477,6 +481,10 @@ RubyLLM.configure do |config|
   config.gpustack_api_base = String
   config.gpustack_api_key = String
 
+  # MiniMax
+  config.minimax_api_key = String
+  config.minimax_api_base = String
+
   # Mistral
   config.mistral_api_key = String
 
diff --git a/lib/ruby_llm.rb b/lib/ruby_llm.rb
index 87bc94c9d..4a22874df 100644
--- a/lib/ruby_llm.rb
+++ b/lib/ruby_llm.rb
@@ -22,6 +22,7 @@
   'deepseek' => 'DeepSeek',
   'gpustack' => 'GPUStack',
   'llm' => 'LLM',
+  'minimax' => 'MiniMax',
   'mistral' => 'Mistral',
   'openai' => 'OpenAI',
   'openrouter' => 'OpenRouter',
@@ -99,6 +100,7 @@ def logger
 RubyLLM::Provider.register :deepseek, RubyLLM::Providers::DeepSeek
 RubyLLM::Provider.register :gemini, RubyLLM::Providers::Gemini
 RubyLLM::Provider.register :gpustack, RubyLLM::Providers::GPUStack
+RubyLLM::Provider.register :minimax, RubyLLM::Providers::MiniMax
 RubyLLM::Provider.register :mistral, RubyLLM::Providers::Mistral
 RubyLLM::Provider.register :ollama, RubyLLM::Providers::Ollama
 RubyLLM::Provider.register :openai, RubyLLM::Providers::OpenAI
diff --git a/lib/ruby_llm/providers/minimax.rb b/lib/ruby_llm/providers/minimax.rb
new file mode 100644
index 000000000..91405317b
--- /dev/null
+++ b/lib/ruby_llm/providers/minimax.rb
@@ -0,0 +1,41 @@
+# frozen_string_literal: true
+
+module RubyLLM
+  module Providers
+    # MiniMax API integration.
+    # MiniMax provides an OpenAI-compatible chat completions API at https://api.minimax.io/v1
+    class MiniMax < OpenAI
+      include MiniMax::Chat
+      include MiniMax::Models
+
+      def api_base
+        @config.minimax_api_base || 'https://api.minimax.io/v1'
+      end
+
+      def headers
+        {
+          'Authorization' => "Bearer #{@config.minimax_api_key}",
+          'Content-Type' => 'application/json'
+        }
+      end
+
+      def maybe_normalize_temperature(temperature, _model)
+        MiniMax::Temperature.normalize(temperature)
+      end
+
+      class << self
+        def capabilities
+          MiniMax::Capabilities
+        end
+
+        def configuration_options
+          %i[minimax_api_key minimax_api_base]
+        end
+
+        def configuration_requirements
+          %i[minimax_api_key]
+        end
+      end
+    end
+  end
+end
diff --git a/lib/ruby_llm/providers/minimax/capabilities.rb b/lib/ruby_llm/providers/minimax/capabilities.rb
new file mode 100644
index 000000000..a7a8bc173
--- /dev/null
+++ b/lib/ruby_llm/providers/minimax/capabilities.rb
@@ -0,0 +1,129 @@
+# frozen_string_literal: true
+
+module RubyLLM
+  module Providers
+    class MiniMax
+      # Determines capabilities and pricing for MiniMax models
+      module Capabilities
+        module_function
+
+        def context_window_for(model_id)
+          case model_id
+          when /M2\.7/ then 1_000_000
+          when /M2\.5/ then 204_000
+          else 204_000
+          end
+        end
+
+        def max_tokens_for(model_id)
+          case model_id
+          when /M2\.7/ then 16_384
+          when /M2\.5/ then 16_384
+          else 8_192
+          end
+        end
+
+        def input_price_for(model_id)
+          PRICES.dig(model_family(model_id), :input) || default_input_price
+        end
+
+        def output_price_for(model_id)
+          PRICES.dig(model_family(model_id), :output) || default_output_price
+        end
+
+        def supports_vision?(_model_id)
+          false
+        end
+
+        def supports_functions?(model_id)
+          model_id.match?(/M2\.[57]/)
+        end
+
+        def supports_tool_choice?(_model_id)
+          true
+        end
+
+        def supports_tool_parallel_control?(_model_id)
+          false
+        end
+
+        def supports_json_mode?(model_id)
+          model_id.match?(/M2\.[57]/)
+        end
+
+        def format_display_name(model_id)
+          model_id
+        end
+
+        def model_type(_model_id)
+          'chat'
+        end
+
+        def model_family(model_id)
+          case model_id
+          when /M2\.7-highspeed/ then :m2_7_highspeed
+          when /M2\.7/ then :m2_7
+          when /M2\.5-highspeed/ then :m2_5_highspeed
+          when /M2\.5/ then :m2_5
+          else :default
+          end
+        end
+
+        PRICES = {
+          m2_7: {
+            input: 0.10,
+            output: 0.10
+          },
+          m2_7_highspeed: {
+            input: 0.07,
+            output: 0.07
+          },
+          m2_5: {
+            input: 0.10,
+            output: 0.10
+          },
+          m2_5_highspeed: {
+            input: 0.07,
+            output: 0.07
+          }
+        }.freeze
+
+        def default_input_price
+          0.10
+        end
+
+        def default_output_price
+          0.10
+        end
+
+        def modalities_for(_model_id)
+          {
+            input: ['text'],
+            output: ['text']
+          }
+        end
+
+        def capabilities_for(model_id)
+          capabilities = ['streaming']
+          capabilities << 'function_calling' if supports_functions?(model_id)
+          capabilities << 'json_mode' if supports_json_mode?(model_id)
+          capabilities
+        end
+
+        def pricing_for(model_id)
+          family = model_family(model_id)
+          prices = PRICES.fetch(family, { input: default_input_price, output: default_output_price })
+
+          {
+            text_tokens: {
+              standard: {
+                input_per_million: prices[:input],
+                output_per_million: prices[:output]
+              }
+            }
+          }
+        end
+      end
+    end
+  end
+end
diff --git a/lib/ruby_llm/providers/minimax/chat.rb b/lib/ruby_llm/providers/minimax/chat.rb
new file mode 100644
index 000000000..2ab0baf95
--- /dev/null
+++ b/lib/ruby_llm/providers/minimax/chat.rb
@@ -0,0 +1,16 @@
+# frozen_string_literal: true
+
+module RubyLLM
+  module Providers
+    class MiniMax
+      # Chat methods of the MiniMax API integration
+      module Chat
+        module_function
+
+        def format_role(role)
+          role.to_s
+        end
+      end
+    end
+  end
+end
diff --git a/lib/ruby_llm/providers/minimax/models.rb b/lib/ruby_llm/providers/minimax/models.rb
new file mode 100644
index 000000000..510fcc7cd
--- /dev/null
+++ b/lib/ruby_llm/providers/minimax/models.rb
@@ -0,0 +1,49 @@
+# frozen_string_literal: true
+
+module RubyLLM
+  module Providers
+    class MiniMax
+      # Models methods of the MiniMax API integration.
+      # MiniMax does not provide a /v1/models endpoint,
+      # so models are statically defined.
+      module Models
+        def list_models(**)
+          slug = 'minimax'
+          capabilities = MiniMax::Capabilities
+          parse_list_models_response(nil, slug, capabilities)
+        end
+
+        def parse_list_models_response(_response, slug, capabilities)
+          model_ids.map do |model_id|
+            create_model_info(model_id, slug, capabilities)
+          end
+        end
+
+        def model_ids
+          %w[
+            MiniMax-M2.7
+            MiniMax-M2.7-highspeed
+            MiniMax-M2.5
+            MiniMax-M2.5-highspeed
+          ]
+        end
+
+        def create_model_info(model_id, slug, capabilities)
+          Model::Info.new(
+            id: model_id,
+            name: capabilities.format_display_name(model_id),
+            provider: slug,
+            family: capabilities.model_family(model_id).to_s,
+            created_at: Time.now,
+            context_window: capabilities.context_window_for(model_id),
+            max_output_tokens: capabilities.max_tokens_for(model_id),
+            modalities: capabilities.modalities_for(model_id),
+            capabilities: capabilities.capabilities_for(model_id),
+            pricing: capabilities.pricing_for(model_id),
+            metadata: {}
+          )
+        end
+      end
+    end
+  end
+end
diff --git a/lib/ruby_llm/providers/minimax/temperature.rb b/lib/ruby_llm/providers/minimax/temperature.rb
new file mode 100644
index 000000000..b0fb5f44f
--- /dev/null
+++ b/lib/ruby_llm/providers/minimax/temperature.rb
@@ -0,0 +1,23 @@
+# frozen_string_literal: true
+
+module RubyLLM
+  module Providers
+    class MiniMax
+      # Normalizes temperature for MiniMax models.
+      # MiniMax accepts temperature in the range [0.0, 1.0].
+      module Temperature
+        module_function
+
+        def normalize(temperature)
+          return temperature if temperature.nil?
+
+          clamped = temperature.to_f.clamp(0.0, 1.0)
+          return clamped if (clamped - temperature.to_f).abs <= Float::EPSILON
+
+          RubyLLM.logger.debug { "MiniMax requires temperature in [0.0, 1.0], clamping #{temperature} to #{clamped}" }
+          clamped
+        end
+      end
+    end
+  end
+end
diff --git a/spec/ruby_llm/providers/minimax/capabilities_spec.rb b/spec/ruby_llm/providers/minimax/capabilities_spec.rb
new file mode 100644
index 000000000..9b53ab99b
--- /dev/null
+++ b/spec/ruby_llm/providers/minimax/capabilities_spec.rb
@@ -0,0 +1,119 @@
+# frozen_string_literal: true
+
+require 'spec_helper'
+
+RSpec.describe RubyLLM::Providers::MiniMax::Capabilities do
+  describe '.context_window_for' do
+    it 'returns 1_000_000 for M2.7 models' do
+      expect(described_class.context_window_for('MiniMax-M2.7')).to eq(1_000_000)
+    end
+
+    it 'returns 1_000_000 for M2.7-highspeed models' do
+      expect(described_class.context_window_for('MiniMax-M2.7-highspeed')).to eq(1_000_000)
+    end
+
+    it 'returns 204_000 for M2.5 models' do
+      expect(described_class.context_window_for('MiniMax-M2.5')).to eq(204_000)
+    end
+
+    it 'returns 204_000 for M2.5-highspeed models' do
+      expect(described_class.context_window_for('MiniMax-M2.5-highspeed')).to eq(204_000)
+    end
+  end
+
+  describe '.max_tokens_for' do
+    it 'returns 16_384 for M2.7 models' do
+      expect(described_class.max_tokens_for('MiniMax-M2.7')).to eq(16_384)
+    end
+
+    it 'returns 16_384 for M2.5 models' do
+      expect(described_class.max_tokens_for('MiniMax-M2.5')).to eq(16_384)
+    end
+  end
+
+  describe '.model_family' do
+    it 'returns :m2_7 for M2.7' do
+      expect(described_class.model_family('MiniMax-M2.7')).to eq(:m2_7)
+    end
+
+    it 'returns :m2_7_highspeed for M2.7-highspeed' do
+      expect(described_class.model_family('MiniMax-M2.7-highspeed')).to eq(:m2_7_highspeed)
+    end
+
+    it 'returns :m2_5 for M2.5' do
+      expect(described_class.model_family('MiniMax-M2.5')).to eq(:m2_5)
+    end
+
+    it 'returns :m2_5_highspeed for M2.5-highspeed' do
+      expect(described_class.model_family('MiniMax-M2.5-highspeed')).to eq(:m2_5_highspeed)
+    end
+  end
+
+  describe '.supports_functions?' do
+    it 'returns true for M2.7 models' do
+      expect(described_class.supports_functions?('MiniMax-M2.7')).to be true
+    end
+
+    it 'returns true for M2.5 models' do
+      expect(described_class.supports_functions?('MiniMax-M2.5')).to be true
+    end
+  end
+
+  describe '.supports_json_mode?' do
+    it 'returns true for M2.7 models' do
+      expect(described_class.supports_json_mode?('MiniMax-M2.7')).to be true
+    end
+
+    it 'returns true for M2.5 models' do
+      expect(described_class.supports_json_mode?('MiniMax-M2.5')).to be true
+    end
+  end
+
+  describe '.supports_vision?' do
+    it 'returns false for all models' do
+      expect(described_class.supports_vision?('MiniMax-M2.7')).to be false
+    end
+  end
+
+  describe '.model_type' do
+    it 'returns chat for all models' do
+      expect(described_class.model_type('MiniMax-M2.7')).to eq('chat')
+    end
+  end
+
+  describe '.format_display_name' do
+    it 'returns the model id as-is' do
+      expect(described_class.format_display_name('MiniMax-M2.7')).to eq('MiniMax-M2.7')
+    end
+  end
+
+  describe '.capabilities_for' do
+    it 'includes streaming for all models' do
+      expect(described_class.capabilities_for('MiniMax-M2.7')).to include('streaming')
+    end
+
+    it 'includes function_calling for M2.7' do
+      expect(described_class.capabilities_for('MiniMax-M2.7')).to include('function_calling')
+    end
+
+    it 'includes json_mode for M2.7' do
+      expect(described_class.capabilities_for('MiniMax-M2.7')).to include('json_mode')
+    end
+  end
+
+  describe '.pricing_for' do
+    it 'returns pricing hash for M2.7' do
+      pricing = described_class.pricing_for('MiniMax-M2.7')
+      expect(pricing).to have_key(:text_tokens)
+      expect(pricing[:text_tokens][:standard]).to include(:input_per_million, :output_per_million)
+    end
+  end
+
+  describe '.modalities_for' do
+    it 'returns text input and output' do
+      modalities = described_class.modalities_for('MiniMax-M2.7')
+      expect(modalities[:input]).to eq(['text'])
+      expect(modalities[:output]).to eq(['text'])
+    end
+  end
+end
diff --git a/spec/ruby_llm/providers/minimax/chat_spec.rb b/spec/ruby_llm/providers/minimax/chat_spec.rb
new file mode 100644
index 000000000..b6ece1dd3
--- /dev/null
+++ b/spec/ruby_llm/providers/minimax/chat_spec.rb
@@ -0,0 +1,19 @@
+# frozen_string_literal: true
+
+require 'spec_helper'
+
+RSpec.describe RubyLLM::Providers::MiniMax::Chat do
+  describe '.format_role' do
+    it 'converts symbol roles to strings' do
+      expect(described_class.format_role(:user)).to eq('user')
+    end
+
+    it 'passes through string roles' do
+      expect(described_class.format_role('assistant')).to eq('assistant')
+    end
+
+    it 'converts system role to string' do
+      expect(described_class.format_role(:system)).to eq('system')
+    end
+  end
+end
diff --git a/spec/ruby_llm/providers/minimax/models_spec.rb b/spec/ruby_llm/providers/minimax/models_spec.rb
new file mode 100644
index 000000000..c2bca2ecf
--- /dev/null
+++ b/spec/ruby_llm/providers/minimax/models_spec.rb
@@ -0,0 +1,40 @@
+# frozen_string_literal: true
+
+require 'spec_helper'
+
+RSpec.describe RubyLLM::Providers::MiniMax::Models do
+  include_context 'with configured RubyLLM'
+
+  let(:provider) { RubyLLM::Providers::MiniMax.new(RubyLLM.config) }
+
+  describe '#list_models' do
+    it 'returns statically defined MiniMax models' do
+      models = provider.list_models
+      model_ids = models.map(&:id)
+
+      expect(model_ids).to include('MiniMax-M2.7', 'MiniMax-M2.7-highspeed',
+                                   'MiniMax-M2.5', 'MiniMax-M2.5-highspeed')
+    end
+
+    it 'returns 4 models' do
+      expect(provider.list_models.size).to eq(4)
+    end
+
+    it 'sets the provider slug to minimax' do
+      models = provider.list_models
+      expect(models).to all(have_attributes(provider: 'minimax'))
+    end
+
+    it 'sets correct context window for M2.7' do
+      models = provider.list_models
+      m27 = models.find { |m| m.id == 'MiniMax-M2.7' }
+      expect(m27.context_window).to eq(1_000_000)
+    end
+
+    it 'sets correct context window for M2.5-highspeed' do
+      models = provider.list_models
+      m25hs = models.find { |m| m.id == 'MiniMax-M2.5-highspeed' }
+      expect(m25hs.context_window).to eq(204_000)
+    end
+  end
+end
diff --git a/spec/ruby_llm/providers/minimax/temperature_spec.rb b/spec/ruby_llm/providers/minimax/temperature_spec.rb
new file mode 100644
index 000000000..533708a48
--- /dev/null
+++ b/spec/ruby_llm/providers/minimax/temperature_spec.rb
@@ -0,0 +1,35 @@
+# frozen_string_literal: true
+
+require 'spec_helper'
+
+RSpec.describe RubyLLM::Providers::MiniMax::Temperature do
+  describe '.normalize' do
+    it 'returns nil when temperature is nil' do
+      expect(described_class.normalize(nil)).to be_nil
+    end
+
+    it 'passes through valid temperature values' do
+      expect(described_class.normalize(0.5)).to eq(0.5)
+    end
+
+    it 'accepts temperature 0.0' do
+      expect(described_class.normalize(0.0)).to eq(0.0)
+    end
+
+    it 'accepts temperature 1.0' do
+      expect(described_class.normalize(1.0)).to eq(1.0)
+    end
+
+    it 'clamps temperature above 1.0 to 1.0' do
+      expect(described_class.normalize(1.5)).to eq(1.0)
+    end
+
+    it 'clamps temperature below 0.0 to 0.0' do
+      expect(described_class.normalize(-0.5)).to eq(0.0)
+    end
+
+    it 'clamps temperature 2.0 to 1.0' do
+      expect(described_class.normalize(2.0)).to eq(1.0)
+    end
+  end
+end
diff --git a/spec/ruby_llm/providers/minimax_spec.rb b/spec/ruby_llm/providers/minimax_spec.rb
new file mode 100644
index 000000000..f8fed7088
--- /dev/null
+++ b/spec/ruby_llm/providers/minimax_spec.rb
@@ -0,0 +1,75 @@
+# frozen_string_literal: true
+
+require 'spec_helper'
+
+RSpec.describe RubyLLM::Providers::MiniMax do
+  subject(:provider) { described_class.new(config) }
+
+  let(:config) do
+    instance_double(
+      RubyLLM::Configuration,
+      request_timeout: 300,
+      max_retries: 3,
+      retry_interval: 0.1,
+      retry_interval_randomness: 0.5,
+      retry_backoff_factor: 2,
+      http_proxy: nil,
+      minimax_api_key: 'test-key',
+      minimax_api_base: minimax_api_base
+    )
+  end
+
+  describe '#api_base' do
+    context 'when minimax_api_base is not set' do
+      let(:minimax_api_base) { nil }
+
+      it 'returns the default MiniMax API URL' do
+        expect(provider.api_base).to eq('https://api.minimax.io/v1')
+      end
+    end
+
+    context 'when minimax_api_base is set' do
+      let(:minimax_api_base) { 'https://custom-minimax-endpoint.example.com' }
+
+      it 'returns the custom API URL' do
+        expect(provider.api_base).to eq('https://custom-minimax-endpoint.example.com')
+      end
+    end
+  end
+
+  describe '#headers' do
+    let(:minimax_api_base) { nil }
+
+    it 'returns Authorization header with API key' do
+      expect(provider.headers).to include('Authorization' => 'Bearer test-key')
+    end
+
+    it 'returns Content-Type header' do
+      expect(provider.headers).to include('Content-Type' => 'application/json')
+    end
+  end
+
+  describe '.configuration_options' do
+    it 'includes minimax_api_key and minimax_api_base' do
+      expect(described_class.configuration_options).to contain_exactly(:minimax_api_key, :minimax_api_base)
+    end
+  end
+
+  describe '.configuration_requirements' do
+    it 'requires minimax_api_key' do
+      expect(described_class.configuration_requirements).to contain_exactly(:minimax_api_key)
+    end
+  end
+
+  describe '.capabilities' do
+    it 'returns MiniMax::Capabilities module' do
+      expect(described_class.capabilities).to eq(RubyLLM::Providers::MiniMax::Capabilities)
+    end
+  end
+
+  describe '.slug' do
+    it 'returns minimax' do
+      expect(described_class.slug).to eq('minimax')
+    end
+  end
+end
diff --git a/spec/support/models_to_test.rb b/spec/support/models_to_test.rb
index 29e90a007..d89b099ee 100644
--- a/spec/support/models_to_test.rb
+++ b/spec/support/models_to_test.rb
@@ -14,6 +14,7 @@ def filter_local_providers(models)
   { provider: :deepseek, model: 'deepseek-chat' },
   { provider: :gemini, model: 'gemini-2.5-flash' },
   { provider: :gpustack, model: 'qwen3' },
+  { provider: :minimax, model: 'MiniMax-M2.7' },
   { provider: :mistral, model: 'mistral-small-latest' },
   { provider: :ollama, model: 'qwen3' },
   { provider: :openai, model: 'gpt-5-nano' },
diff --git a/spec/support/rubyllm_configuration.rb b/spec/support/rubyllm_configuration.rb
index 6b3cede88..fcff717d3 100644
--- a/spec/support/rubyllm_configuration.rb
+++ b/spec/support/rubyllm_configuration.rb
@@ -25,6 +25,7 @@
       config.gpustack_api_key = ENV.fetch('GPUSTACK_API_KEY', nil)
       # Disable retries in tests for deterministic, fast failures.
       config.max_retries = 0
+      config.minimax_api_key = ENV.fetch('MINIMAX_API_KEY', 'test')
       config.mistral_api_key = ENV.fetch('MISTRAL_API_KEY', 'test')
       config.model_registry_class = 'Model'
       config.ollama_api_base = ENV.fetch('OLLAMA_API_BASE', 'http://localhost:11434/v1')