支持的平台

对比 AI 提供商与模型

接入 11 家主流 AI 提供商，从 92 款模型中选择，涵盖 ChatGPT、Claude、Gemini、Grok、Kimi 和 MiniMax。查看可用选项、对比差异，无需重建产品即可切换提供商。.

AI 提供商
11

团队熟悉的主流平台：OpenAI、Anthropic、Google、xAI、Cohere、Mistral、DeepSeek、Qwen、Groq、Kimi 和 MiniMax。.

可用模型
92

此处列出的模型均可在 ModelRiver 上直接使用，无需您额外配置。.

一次接入
1 次配置

团队只需接入一次。后续更换模型或提供商，无需推倒重来。.

OpenAI

GPT-5.6 Sol/Terra/Luna, GPT-5.5/5.4, GPT-4.1/4o, GPT-5.3 Codex, and O-series reasoning models

29 models
Frontier, mini, codex, and reasoning tiers
gpt-5.6-sol
GPT-5.6 Sol - flagship frontier model for complex reasoning and coding
Chat Completion Text Vision Streaming Structured output Tools 400K ctx 128K out
400K context · 128K max out · Released 2026
$5 in
$30 out
gpt-5.6-terra
GPT-5.6 Terra - balances intelligence and cost for everyday professional work
Chat Completion Text Vision Streaming Structured output Tools 400K ctx 128K out
400K context · 128K max out · Released 2026
$2.50 in
$15 out
gpt-5.6-luna
GPT-5.6 Luna - fast low-cost model for high-volume workloads
Chat Completion Text Vision Streaming Structured output Tools 400K ctx 128K out
400K context · 128K max out · Released 2026
$1 in
$6 out
gpt-5.5
Previous frontier GPT model for coding and professional work
Chat Completion Text Vision Streaming Structured output Tools 400K ctx 128K out
400K context · 128K max out · Released 2026
$5 in
$30 out
gpt-5.5-pro
Highest-compute GPT-5.5 variant
Chat Completion Text Vision Streaming Structured output Tools 400K ctx 128K out
400K context · 128K max out · Released 2026
$30 in
$180 out
gpt-5.4
Production workhorse GPT model for professional work
Chat Completion Text Vision Streaming Structured output Tools 1M ctx 128K out
1M context · 128K max out · Released 2026
$2.50 in
$15 out
gpt-5.4-mini
Strong mini GPT model for coding and subagents
Chat Completion Text Vision Streaming Structured output Tools 400K ctx 128K out
400K context · 128K max out · Released 2026
$0.75 in
$4.50 out
gpt-5.4-nano
Smallest current GPT-5.4 variant
Chat Completion Text Vision Streaming Structured output Tools 400K ctx 128K out
400K context · 128K max out · Released 2026
$0.2 in
$1.25 out
gpt-5.4-pro
Highest-precision GPT-5.4 variant
Chat Completion Text Vision Streaming Structured output Tools 1M ctx 128K out
1M context · 128K max out · Released 2026
$30 in
$180 out
gpt-5.3-codex
Latest Codex model for agentic coding
Chat Completion Text Vision Streaming Structured output Tools 400K ctx 128K out
400K context · 128K max out · Released 2026
$1.75 in
$14 out
gpt-5.2
Previous frontier GPT model with configurable reasoning
Chat Completion Text Vision Streaming Structured output Tools 400K ctx 128K out
400K context · 128K max out · Released 2025
$1.75 in
$14 out
gpt-5.2-pro
Previous pro GPT model for professional work
Chat Completion Text Vision Streaming Structured output Tools 400K ctx 128K out
400K context · 128K max out · Released 2025
$21 in
$168 out
gpt-5.1
Flagship GPT model for coding and agentic tasks
Chat Completion Text Vision Streaming Structured output Tools 400K ctx 128K out
400K context · 128K max out · Released 2025
$1.25 in
$10 out
gpt-5
Previous GPT-5 reasoning model
Chat Completion Text Vision Streaming Structured output Tools 400K ctx 128K out
400K context · 128K max out · Released 2025
$1.25 in
$10 out
gpt-5-pro
Higher-compute GPT-5 variant
Chat Completion Text Vision Streaming Structured output Tools 400K ctx 128K out
400K context · 128K max out · Released 2025
$15 in
$120 out
gpt-5-mini
Faster, cost-efficient GPT-5 variant
Chat Completion Text Vision Streaming Structured output Tools 400K ctx 128K out
400K context · 128K max out · Released 2025
$0.25 in
$2 out
gpt-5-nano
Fastest and cheapest GPT-5 variant
Chat Completion Text Vision Streaming Structured output Tools 400K ctx 128K out
400K context · 128K max out · Released 2025
$0.05 in
$0.4 out
o3-pro
Higher-compute o3 reasoning model
Chat Completion Text Vision Streaming Structured output Tools 200K ctx 100K out
200K context · 100K max out · Released 2025
$20 in
$80 out
o3
Powerful reasoning model across text, code, and vision
Chat Completion Text Vision Streaming Structured output Tools 200K ctx 100K out
200K context · 100K max out · Released 2025
$2 in
$8 out
o4-mini
Fast, cost-efficient reasoning model
Chat Completion Text Vision Streaming Structured output Tools 200K ctx 100K out
200K context · 100K max out · Released 2025
$1.10 in
$4.40 out
gpt-4.1
Strong non-reasoning model for instructions and tools
Chat Completion Text Vision Streaming Structured output Tools 1.0M ctx 32.8K out
1.0M context · 32.8K max out · Released 2025
$2 in
$8 out
gpt-4.1-mini
Efficient GPT-4.1 variant
Chat Completion Text Vision Streaming Structured output Tools 1.0M ctx 32.8K out
1.0M context · 32.8K max out · Released 2025
$0.4 in
$1.60 out
gpt-4.1-nano
Lightweight GPT-4.1 model
Chat Completion Text Vision Streaming Structured output Tools 1.0M ctx 32.8K out
1.0M context · 32.8K max out · Released 2025
$0.1 in
$0.4 out
o3-mini
Small reasoning model alternative to o3
Chat Completion Text Vision Streaming Structured output Tools 200K ctx 100K out
200K context · 100K max out · Released 2025
$1.10 in
$4.40 out
o1-pro
Higher-compute legacy reasoning model still available in API
Chat Completion Text Vision Streaming Structured output Tools 200K ctx 100K out
200K context · 100K max out · Released 2024
$150 in
$600 out
gpt-4o-mini
Low-cost multimodal small model
Chat Completion Text Vision Streaming Structured output Tools 128K ctx 16.4K out
128K context · 16.4K max out · Released 2024
$0.15 in
$0.6 out
gpt-4o
Widely used multimodal flagship model
Chat Completion Text Vision Streaming Structured output Tools 128K ctx 16.4K out
128K context · 16.4K max out · Released 2024
$2.50 in
$10 out
text-embedding-3-small
Latest OpenAI small embedding model — efficient default for RAG
Embedding Text 8.2K ctx
8.2K context · Released 2024
$0.02 in
$0 out
text-embedding-3-large
Latest OpenAI large embedding model with highest accuracy
Embedding Text 8.2K ctx
8.2K context · Released 2024
$0.13 in
$0 out

Anthropic

Claude Fable 5, Opus 4.8, Sonnet 5, and Haiku 4.5

claude-fable-5
Most capable Claude model for long-horizon agentic work
Chat Completion Text Vision Streaming Structured output Tools 1M ctx 64K out
1M context · 64K max out · Released 2026
$10 in
$50 out
claude-opus-4-8
Latest Claude Opus model for complex agents
Chat Completion Text Vision Streaming Structured output Tools 200K ctx 32K out
200K context · 32K max out · Released 2026
$5 in
$25 out
claude-sonnet-5
Latest Claude Sonnet - best speed/intelligence balance (intro pricing)
Chat Completion Text Vision Streaming Structured output Tools 200K ctx 64K out
200K context · 64K max out · Released 2026
$2 in
$10 out
claude-haiku-4-5-20251001
Claude Haiku 4.5 - low latency and cost-efficient current Haiku
Chat Completion Text Vision Streaming Structured output Tools 200K ctx 64K out
200K context · 64K max out · Released 2025
$1 in
$5 out

Google

Gemini 3.6 Flash, 3.5 Flash/Flash-Lite, 3.1 Pro/Flash-Lite, and Gemini 2.5 family

gemini-3.6-flash
Gemini 3.6 Flash - latest GA Flash for agentic and multimodal work
Chat Completion Text Vision Streaming Structured output Tools 1.0M ctx 65.5K out
1.0M context · 65.5K max out · Released 2026
$1.50 in
$7.50 out
gemini-3.5-flash-lite
Gemini 3.5 Flash-Lite - fastest low-cost 3.5 model for high throughput
Chat Completion Text Vision Streaming Structured output Tools 1.0M ctx 65.5K out
1.0M context · 65.5K max out · Released 2026
$0.3 in
$2.50 out
gemini-3.5-flash
Gemini 3.5 Flash - frontier speed with strong search and grounding
Chat Completion Text Vision Streaming Structured output Tools 1.0M ctx 65.5K out
1.0M context · 65.5K max out · Released 2026
$1.50 in
$9 out
gemini-embedding-2
Latest Gemini embedding model (GA Apr 2026) for text and multimodal RAG
Embedding Text 8.2K ctx
8.2K context · Released 2026
$0.15 in
$0 out
gemini-3.1-pro-preview
Gemini 3.1 Pro Preview - latest multimodal reasoning model
Chat Completion Text Vision Streaming Structured output Tools 1.0M ctx 65.5K out
1.0M context · 65.5K max out · Released 2026
$2 in
$12 out
gemini-3.1-flash-lite
Gemini 3.1 Flash-Lite - cost-efficient high-volume Gemini 3.1 model
Chat Completion Text Vision Streaming Structured output Tools 1.0M ctx 65.5K out
1.0M context · 65.5K max out · Released 2026
$0.25 in
$1.50 out
gemini-3-flash-preview
Gemini 3 Flash Preview - Pro-level intelligence at Flash pricing
Chat Completion Text Vision Streaming Structured output Tools 1.0M ctx 65.5K out
1.0M context · 65.5K max out · Released 2025
$0.5 in
$3 out
gemini-2.5-flash-lite
Gemini 2.5 Flash-Lite - ultra cost-efficient model
Chat Completion Text Vision Streaming Structured output Tools 1.0M ctx 65.5K out
1.0M context · 65.5K max out · Released 2025
$0.1 in
$0.4 out
gemini-2.5-flash
Gemini 2.5 Flash - fast high-throughput model
Chat Completion Text Vision Streaming Structured output Tools 1.0M ctx 65.5K out
1.0M context · 65.5K max out · Released 2025
$0.3 in
$2.50 out
gemini-2.5-pro
Gemini 2.5 Pro - high-capability multipurpose model
Chat Completion Text Vision Streaming Structured output Tools 1.0M ctx 65.5K out
1.0M context · 65.5K max out · Released 2025
$1.25 in
$10 out

xAI

Grok 4.5 - xAI's flagship model for coding, chat, and agents

grok-4.5
Grok 4.5 - current flagship for coding, chat, and agents
Chat Completion Text Streaming Structured output Tools 256K ctx 64K out
256K context · 64K max out · Released 2026
$2 in
$6 out

Cohere

Command A+, Command A, and current Command R family models

command-a-plus-05-2026
Command A+ - latest enterprise RAG and agentic model (May 2026)
Chat Completion Text Streaming Structured output Tools 256K ctx 8.2K out
256K context · 8.2K max out · Released 2026
$2.50 in
$10 out
command-a-reasoning-08-2025
Command A Reasoning - reasoning-capable Command A variant
Chat Completion Text Streaming Structured output Tools 256K ctx 32K out
256K context · 32K max out · Released 2025
$2.50 in
$10 out
command-a-translate-08-2025
Command A Translate - state-of-the-art translation model
Chat Completion Text Streaming Structured output Tools 128K ctx 8.2K out
128K context · 8.2K max out · Released 2025
$2.50 in
$10 out
command-a-vision-07-2025
Command A Vision - multimodal Command A model for image inputs
Chat Completion Text Vision Streaming Structured output Tools 128K ctx 8.2K out
128K context · 8.2K max out · Released 2025
$2.50 in
$10 out
command-a-03-2025
Command A - advanced instruction-following model (March 2025)
Chat Completion Text Streaming Structured output Tools 256K ctx 8.2K out
256K context · 8.2K max out · Released 2025
$2.50 in
$10 out
embed-v4.0
Latest Cohere embedding model for search, RAG, and multimodal retrieval
Embedding Text 128K ctx
128K context · Released 2025
$0.12 in
$0 out
command-r7b-12-2024
Command R7B - fastest low-cost Command model
Chat Completion Text Streaming Structured output Tools 128K ctx 4.1K out
128K context · 4.1K max out · Released 2024
$0.0375 in
$0.15 out
command-r-plus-08-2024
Command R+ - strong long-context agent model
Chat Completion Text Streaming Structured output Tools 128K ctx 4.1K out
128K context · 4.1K max out · Released 2024
$2.50 in
$10 out
command-r-08-2024
Command R - lower-cost long-context model
Chat Completion Text Streaming Structured output Tools 128K ctx 4.1K out
128K context · 4.1K max out · Released 2024
$0.15 in
$0.6 out

Mistral AI

Mistral Large/Medium/Small, Devstral, Magistral, Ministral, and Codestral

mistral-small-latest
Mistral Small 4 - fast and efficient multimodal model
Chat Completion Text Streaming Structured output Tools 128K ctx 32K out
128K context · 32K max out · Released 2026
$0.1 in
$0.3 out
devstral-medium-latest
Devstral 2 - agentic coding model
Chat Completion Text Streaming Structured output Tools 256K ctx 64K out
256K context · 64K max out · Released 2025
$0.4 in
$2 out
devstral-small-latest
Devstral Small 2 - lightweight coding agent model
Chat Completion Text Streaming Structured output Tools 128K ctx 32K out
128K context · 32K max out · Released 2025
$0.1 in
$0.3 out
ministral-14b-latest
Ministral 3 14B - compact text and vision model
Chat Completion Text Streaming Structured output Tools 256K ctx 32K out
256K context · 32K max out · Released 2025
$0.2 in
$0.2 out
ministral-8b-latest
Ministral 3 8B - efficient edge-capable model
Chat Completion Text Streaming Structured output Tools 256K ctx 32K out
256K context · 32K max out · Released 2025
$0.15 in
$0.15 out
ministral-3b-latest
Ministral 3 3B - smallest Ministral 3 model
Chat Completion Text Streaming Structured output Tools 256K ctx 32K out
256K context · 32K max out · Released 2025
$0.1 in
$0.1 out
mistral-large-latest
Mistral Large 3 - flagship general-purpose model
Chat Completion Text Streaming Structured output Tools 256K ctx 64K out
256K context · 64K max out · Released 2025
$0.5 in
$1.50 out
mistral-medium-latest
Mistral Medium 3.5 - frontier multimodal model for agents and coding
Chat Completion Text Streaming Structured output Tools 128K ctx 64K out
128K context · 64K max out · Released 2025
$1.50 in
$7.50 out
magistral-medium-latest
Magistral Medium - multimodal reasoning model
Chat Completion Text Streaming Structured output Tools 128K ctx 40K out
128K context · 40K max out · Released 2025
$2 in
$5 out
magistral-small-latest
Magistral Small - lightweight reasoning model
Chat Completion Text Streaming Structured output Tools 128K ctx 40K out
128K context · 40K max out · Released 2025
$0.5 in
$1.50 out
codestral-latest
Codestral - specialized coding model
Chat Completion Text Streaming Structured output Tools 256K ctx 32K out
256K context · 32K max out · Released 2025
$0.3 in
$0.9 out
codestral-embed
Mistral code-focused embedding model
Embedding Text 8.2K ctx
8.2K context · Released 2025
$0.15 in
$0 out
mistral-embed
Mistral embedding model for semantic search and RAG
Embedding Text 8.2K ctx
8.2K context · Released 2023
$0.1 in
$0 out

DeepSeek

DeepSeek V4 Pro and V4 Flash

deepseek-v4-pro
DeepSeek V4 Pro - frontier coding and long-horizon agents
Chat Completion Text Streaming Structured output Tools 1M ctx 64K out
1M context · 64K max out · Released 2026
$0.435 in
$0.87 out
deepseek-v4-flash
DeepSeek V4 Flash - default chat and high-volume workloads
Chat Completion Text Streaming Structured output Tools 1M ctx 64K out
1M context · 64K max out · Released 2026
$0.14 in
$0.28 out

Qwen

Qwen 3.7 Max/Plus, 3.6 Flash, 3.5 Plus/Flash, and Coder (Alibaba Cloud)

qwen3.7-max
Qwen 3.7 Max - latest flagship Qwen model
Chat Completion Text Streaming Structured output Tools 262.1K ctx 65.5K out
262.1K context · 65.5K max out · Released 2026
$2.50 in
$7.50 out
qwen3.7-plus
Qwen 3.7 Plus - latest balanced production model
Chat Completion Text Streaming Structured output Tools 1M ctx 65.5K out
1M context · 65.5K max out · Released 2026
$0.4 in
$1.60 out
qwen3.6-flash
Qwen 3.6 Flash - fast long-context cost-efficient model
Chat Completion Text Streaming Structured output Tools 1M ctx 65.5K out
1M context · 65.5K max out · Released 2026
$0.25 in
$1.50 out
qwen3.5-plus
Qwen 3.5 Plus - balanced production model
Chat Completion Text Streaming Structured output Tools 1M ctx 65.5K out
1M context · 65.5K max out · Released 2026
$0.4 in
$2.40 out
qwen3.5-flash
Qwen 3.5 Flash - fast 1M-context cost-efficient model
Chat Completion Text Streaming Structured output Tools 1M ctx 65.5K out
1M context · 65.5K max out · Released 2026
$0.1 in
$0.4 out
qwen3-coder-plus
Qwen3 Coder Plus - current coding and agent model
Chat Completion Text Streaming Structured output Tools 1M ctx 65.5K out
1M context · 65.5K max out · Released 2025
$1 in
$5 out
qwen3-max
Qwen3 Max - previous flagship Qwen model
Chat Completion Text Streaming Structured output Tools 262.1K ctx 65.5K out
262.1K context · 65.5K max out · Released 2025
$1.20 in
$6 out
qwen-flash
Qwen-Flash - fast and cost-efficient flagship model
Chat Completion Text Streaming Structured output Tools 1M ctx 32.8K out
1M context · 32.8K max out · Released 2025
$0.05 in
$0.4 out
qwen-max
Qwen-Max - flagship model, highest capability
Chat Completion Text Streaming Structured output Tools 131.1K ctx 16.4K out
131.1K context · 16.4K max out · Released 2025
$1.60 in
$6.40 out
qwen-plus
Qwen-Plus - balanced performance and cost
Chat Completion Text Streaming Structured output Tools 131.1K ctx 16.4K out
131.1K context · 16.4K max out · Released 2025
$0.4 in
$1.20 out

Groq

Ultra-fast LPU inference - Llama 3.1/3.3, GPT OSS

openai/gpt-oss-120b
GPT OSS 120B - open model with strong quality/cost balance
Chat Completion Text Streaming Structured output Tools 128K ctx 32.8K out
128K context · 32.8K max out · Released 2025
$0.15 in
$0.6 out
openai/gpt-oss-20b
GPT OSS 20B - very fast open model for lightweight workloads
Chat Completion Text Streaming Structured output Tools 128K ctx 32.8K out
128K context · 32.8K max out · Released 2025
$0.075 in
$0.3 out
llama-3.3-70b-versatile
Llama 3.3 70B - versatile high-capability model
Chat Completion Text Streaming Structured output Tools 128K ctx 32.8K out
128K context · 32.8K max out · Released 2024
$0.59 in
$0.79 out
llama-3.1-8b-instant
Llama 3.1 8B - fast low-latency model
Chat Completion Text Streaming Structured output Tools 128K ctx 8.2K out
128K context · 8.2K max out · Released 2024
$0.05 in
$0.08 out

Kimi

Kimi K3 flagship, K2.7 Code (standard and high-speed), K2.6, and K2.5 models

kimi-k3
Kimi K3 - flagship 2.8T multimodal model with 1M context
Chat Completion Text Streaming Structured output Tools 1.0M ctx 131.1K out
1.0M context · 131.1K max out · Released 2026
$3 in
$15 out
kimi-k2.7-code
Kimi K2.7 Code - latest coding model with 256K context
Chat Completion Text Streaming Structured output Tools 256K ctx 32K out
256K context · 32K max out · Released 2026
$0.95 in
$4 out
kimi-k2.7-code-highspeed
Kimi K2.7 Code Highspeed - the same coding model with higher output throughput
Chat Completion Text Streaming Structured output Tools 256K ctx 32K out
256K context · 32K max out · Released 2026
$1.90 in
$8 out
kimi-k2.6
Kimi K2.6 - flagship multimodal agent and coding model (256K context)
Chat Completion Text Streaming Structured output Tools 256K ctx 32K out
256K context · 32K max out · Released 2026
$0.95 in
$4 out
kimi-k2.5
Kimi K2.5 - multimodal model with thinking and non-thinking modes (256K context)
Chat Completion Text Streaming Structured output Tools 256K ctx 32K out
256K context · 32K max out · Released 2026
$0.6 in
$3 out

MiniMax

MiniMax M3 and M2.7/M2.5 coding/agent models

MiniMax-M3
MiniMax M3 - latest agentic reasoning and coding model (1M context)
Chat Completion Text Streaming Structured output Tools 1M ctx 64K out
1M context · 64K max out · Released 2026
$0.3 in
$1.20 out
MiniMax-M2.7
MiniMax M2.7 - flagship MoE coding and agent model (205K context)
Chat Completion Text Streaming Structured output Tools 205K ctx 128K out
205K context · 128K max out · Released 2026
$0.3 in
$1.20 out
MiniMax-M2.7-highspeed
MiniMax M2.7 Highspeed - same M2.7 quality with higher throughput
Chat Completion Text Streaming Structured output Tools 205K ctx 128K out
205K context · 128K max out · Released 2026
$0.6 in
$2.40 out
MiniMax-M2.5
MiniMax M2.5 - previous flagship coding/agent model, still supported
Chat Completion Text Streaming Structured output Tools 205K ctx 128K out
205K context · 128K max out · Released 2026
$0.3 in
$1.20 out
MiniMax-M2.5-highspeed
MiniMax M2.5 Highspeed - previous M2.5 with higher throughput
Chat Completion Text Streaming Structured output Tools 205K ctx 128K out
205K context · 128K max out · Released 2026
$0.6 in
$2.40 out

一次搭建，随 AI 演变保持灵活

Set up your AI 工作流 once, then move between 提供商, including Kimi and MiniMax, when pricing, performance, or strategy shifts, without starting from scratch.

免费开始 查看文档