新闻

2026年LLM API费用对比——免费在线工具 | TokenCost

新闻 2026-05-12 0 次浏览

API Pricing

Every major LLM provider in one table. Sortable, filterable. Prices per 1 million tokens.

Model ▲▼Provider ▲▼Input / 1M Output / 1M ▲▼Context ▲▼Max OutputSpeed ▲▼Quality ▲▼Value ▲▼
GPT-OSS 120BOpenAI$0.039$0.190131K33K------
GPT-5 NanoOpenAI$0.050$0.400128K16K141 tok/s27536.0
Qwen3.5-9BAlibaba$0.050$0.150262K33K------
Qwen3.5-Omni FlashAlibaba$0.065$0.260262K33K------
Hunyuan HY3 PreviewTencent$0.066$0.260262K33K------
GPT-OSS 20B (Bedrock)OpenAI$0.070$0.30016K16K------
GLM-4.7-flashZhipu$0.070$0.400200K8K------
GPT-OSS 20BOpenAI$0.075$0.300131K33K------
Gemini 2.0 Flash-LiteGoogle$0.075$0.3001.0M8K--15193.3
Llama 4 ScoutMeta$0.080$0.3001.0M16K136 tok/s14168.8
Qwen3-Next-80B-A3B-ThinkingAlibaba$0.098$0.780262K33K------
GPT-4.1 NanoOpenAI$0.100$0.4001.0M33K127 tok/s13130.0
Gemini 2.5 Flash-LiteGoogle$0.100$0.4001.0M66K263 tok/s13127.0
Gemini 2.0 FlashGoogle$0.100$0.4001.0M8K--19185.0
Devstral SmallMistral$0.100$0.300256K33K------
Mistral Small 3.2Mistral$0.100$0.300128K4K147 tok/s10102.0
Gemma 4 26B A4BGoogle$0.130$0.400262K33K------
Gemma 4 31BGoogle$0.140$0.400262K33K------
DeepSeek V4-FlashDeepSeek$0.140$0.2801.0M384K------
GPT-OSS 120B (Bedrock)OpenAI$0.150$0.60016K16K------
GPT-4o MiniOpenAI$0.150$0.600128K16K80 tok/s1384.0
Mistral Small 4Mistral$0.150$0.600256K33K------
Ministral 3 8BMistral$0.150$0.150262K33K------
Pixtral 12BMistral$0.150$0.150128K4K------
Command RCohere$0.150$0.600128K4K------
Command R 08 2024Cohere$0.150$0.600128K4K------
Command R7b 12 2024Cohere$0.150$0.038128K4K------
Llama 3.3 70BMeta$0.180$0.180131K4K------
GPT-5.4 NanoOpenAI$0.200$1.25400K128K------
Grok 4.1 FastxAI$0.200$0.5002.0M16K78 tok/s24118.0
Grok 4.1 Fast ReasoningxAI$0.200$0.5002.0M16K95 tok/s39193.0
Grok Code FastxAI$0.200$1.50256K16K------
Ministral 3 14BMistral$0.200$0.200262K33K------
Jamba 1.5AI21$0.200$0.400256K256K------
Jamba 1.5 MiniAI21$0.200$0.400256K256K------
Jamba 1.5 Mini@001AI21$0.200$0.400256K256K------
Jamba Mini 1.6AI21$0.200$0.400256K256K------
Jamba Mini 1.7AI21$0.200$0.400256K256K------
GPT-5 MiniOpenAI$0.250$2.00400K16K103 tok/s41164.8
Gemini 3.1 Flash-LiteGoogle$0.250$1.501.0M66K317 tok/s34134.0
Claude 3 Haiku 20240307Anthropic$0.250$1.25200K4K------
Llama 4 MaverickMeta$0.270$0.8501.0M16K113 tok/s1868.1
Qwen3.6-PlusAlibaba$0.276$1.651.0M33K------
DeepSeek V3.2 (Chat)DeepSeek$0.280$0.420128K8K--32114.6
DeepSeek V3.2 (Reasoner)DeepSeek$0.280$0.420128K8K------
DeepSeek ReasonerDeepSeek$0.280$0.420131K66K------
Gemini 2.5 FlashGoogle$0.300$2.501.0M66K215 tok/s2168.7
Grok 3 MinixAI$0.300$0.500131K16K------
CodestralMistral$0.300$0.900256K33K------
Qwen 3.5 27BAlibaba$0.300$2.40128K33K------
Nova 2.0 LiteAmazon$0.300$2.501.0M64K228 tok/s1860.0
Nemotron 3 Super 120BNVIDIA$0.300$0.8001.0M33K157 tok/s36120.0
MiniMax M2.5MiniMax$0.300$1.20128K33K94 tok/s42139.7
Command LightCohere$0.300$0.6004K4K------
GPT-4.1 MiniOpenAI$0.400$1.601.0M33K79 tok/s2357.2
Mistral MediumMistral$0.400$2.00131K16K------
DevstralMistral$0.400$2.00256K33K------
Qwen3.5-Omni PlusAlibaba$0.400$4.80262K33K------
Gemini 3 FlashGoogle$0.500$3.001.0M66K198 tok/s3570.0
Gemini 3 Flash ReasoningGoogle$0.500$3.001.0M66K205 tok/s4692.8
Mistral Large 3Mistral$0.500$1.50262K8K56 tok/s2345.6
Magistral SmallMistral$0.500$1.5040K16K------
DeepSeek R1DeepSeek$0.550$2.19128K33K--2749.3
Grok 3 Mini FastxAI$0.600$4.00131K16K------
Qwen 3.5 397BAlibaba$0.600$3.60128K33K------
Kimi K2.5Moonshot$0.600$3.00262K33K47 tok/s4778.0
Kimi K2 ThinkingMoonshot$0.600$2.50262K33K------
GLM-4.7Zhipu$0.600$2.20200K128K------
GPT-5.4 MiniOpenAI$0.750$4.50400K128K------
Gemini 3.1 Flash LiveGoogle$0.750$4.501.0M66K------
Claude Haiku 3.5Anthropic$0.800$4.00200K8K--1923.4
QwQ-PlusAlibaba$0.800$2.40131K16K------
Kimi K2.6Moonshot$0.950$4.00262K33K------
Claude Haiku 4.5Anthropic$1.00$5.00200K8K103 tok/s3131.1
Claude 4.5 Haiku ReasoningAnthropic$1.00$5.00200K8K139 tok/s3737.1
Gemini 3.1 Flash TTSGoogle$1.00$20.0032K32K------
GLM-5Zhipu$1.00$3.20128K33K72 tok/s5049.8
MiMo-V2-ProXiaomi$1.00$3.001.0M131K------
Claude Haiku 4 5 20251001Anthropic$1.00$5.00200K64K------
Claude Haiku 4 5Anthropic$1.00$5.00200K64K------
o4 MiniOpenAI$1.10$4.40200K100K168 tok/s3330.1
o3 MiniOpenAI$1.10$4.40200K100K158 tok/s2623.5
Kimi K2 Thinking TurboMoonshot$1.15$8.00262K33K------
GLM-5 TurboZhipu$1.20$4.00200K128K------
GPT-5.1OpenAI$1.25$10.00400K16K150 tok/s4838.2
GPT-5OpenAI$1.25$10.00400K16K95 tok/s4535.7
GPT-5 MediumOpenAI$1.25$10.00400K16K85 tok/s4233.6
Gemini 2.5 ProGoogle$1.25$10.001.0M66K136 tok/s3527.7
Grok 4.3xAI$1.25$2.501.0M16K------
Nova 2.0 Pro ReasoningAmazon$1.25$10.00128K33K155 tok/s3628.6
GLM-5.1Zhipu$1.40$4.40200K128K------
Mistral Medium 3.5Mistral$1.50$7.50256K33K------
DeepSeek V4-ProDeepSeek$1.74$3.481.0M384K------
GPT-5.2OpenAI$1.75$14.00400K16K79 tok/s5129.3
GPT-5.3 CodexOpenAI$1.75$14.00400K33K98 tok/s5430.6
GPT-4.1OpenAI$2.00$8.001.0M33K123 tok/s2613.2
o3OpenAI$2.00$8.00200K100K113 tok/s3819.2
o4 Mini Deep ResearchOpenAI$2.00$8.00200K100K------
Gemini 3.1 ProGoogle$2.00$12.001.0M66K147 tok/s5728.6
Gemini 3 ProGoogle$2.00$12.001.0M66K141 tok/s4824.2
Grok 4.20xAI$2.00$6.002.0M16K94 tok/s4924.6
Grok 2xAI$2.00$10.00131K16K------
Magistral MediumMistral$2.00$5.0040K16K------
Pixtral LargeMistral$2.00$6.00128K4K------
Jamba 1.5 LargeAI21$2.00$8.00256K256K------
Jamba 1.5 Large@001AI21$2.00$8.00256K256K------
Jamba Large 1.6AI21$2.00$8.00256K256K------
Jamba Large 1.7AI21$2.00$8.00256K256K------
GPT-5.4OpenAI$2.50$15.001.1M128K95 tok/s5722.7
GPT-4oOpenAI$2.50$10.00128K16K149 tok/s176.9
Command ACohere$2.50$10.00128K4K34 tok/s145.4
Command A 03 2025Cohere$2.50$10.00256K8K------
Command R PlusCohere$2.50$10.00128K4K------
Command R Plus 08 2024Cohere$2.50$10.00128K4K------
Claude Sonnet 4.6 AdaptiveAnthropic$3.00$15.00200K64K69 tok/s5217.2
Claude Sonnet 4.6Anthropic$3.00$15.00200K64K51 tok/s4414.8
Claude Sonnet 4.5Anthropic$3.00$15.00200K64K------
Claude Sonnet 4Anthropic$3.00$15.00200K64K57 tok/s4314.2
Claude 3.7 SonnetAnthropic$3.00$15.00200K64K------
Grok 4xAI$3.00$15.002.0M16K46 tok/s4213.8
Grok 3xAI$3.00$15.00131K16K------
Sonar ProPerplexity$3.00$15.00128K4K--155.1
Claude 3 7 Sonnet 20250219Anthropic$3.00$15.00200K64K------
Claude 4 Sonnet 20250514Anthropic$3.00$15.001.0M64K------
Claude Sonnet 4 5Anthropic$3.00$15.00200K64K------
Claude Sonnet 4 5 20250929Anthropic$3.00$15.00200K64K------
Claude Sonnet 4 6Anthropic$3.00$15.001.0M64K------
Claude Sonnet 4 20250514Anthropic$3.00$15.001.0M64K------
GPT-5.5OpenAI$5.00$30.001.1M128K------
Claude Opus 4.7Anthropic$5.00$25.001.0M128K------
Claude Opus 4.6 AdaptiveAnthropic$5.00$25.00200K32K53 tok/s5310.6
Claude Opus 4.6Anthropic$5.00$25.00200K32K44 tok/s479.3
Claude Opus 4.5Anthropic$5.00$25.00200K32K59 tok/s438.6
Grok 3 FastxAI$5.00$25.00131K16K------
MAI-Image-2Microsoft$5.00$33.0032K1K------
Claude Opus 4 5 20251101Anthropic$5.00$25.00200K64K------
Claude Opus 4 5Anthropic$5.00$25.00200K64K------
Claude Opus 4 6Anthropic$5.00$25.001.0M128K------
Claude Opus 4 6 20260205Anthropic$5.00$25.001.0M128K------
Claude Opus 4 7Anthropic$5.00$25.001.0M128K------
Claude Opus 4 7 20260416Anthropic$5.00$25.001.0M128K------
o3 Deep ResearchOpenAI$10.00$40.00200K100K------
o1OpenAI$15.00$60.00200K100K102 tok/s312.1
Claude Opus 4.1Anthropic$15.00$75.00200K32K------
Claude 3 Opus 20240229Anthropic$15.00$75.00200K4K------
Claude 4 Opus 20250514Anthropic$15.00$75.00200K32K------
Claude Opus 4 1Anthropic$15.00$75.00200K32K------
Claude Opus 4 1 20250805Anthropic$15.00$75.00200K32K------
Claude Opus 4 20250514Anthropic$15.00$75.00200K32K------
Voxtral TTSMistral$16.00$< 0.01128K0K------
o3-proOpenAI$20.00$80.00200K100K------
Claude Mythos PreviewAnthropic$25.00$1251.0M32K------
GPT-5.5 ProOpenAI$30.00$1801.1M128K------
GPT-5.4 ProOpenAI$30.00$1801.1M128K------
o1 ProOpenAI$150$600200K100K------

Pricing as of March 2026. Open-source model prices reflect hosted API providers. Always verify with official pages. Value = quality index / input cost per 1M tokens (higher is better).

Speed & quality data by Artificial Analysis

How to Compare LLM API Pricing

  1. 1

    Browse the pricing table

    400+ models are listed with input and output pricing per million tokens, plus context window sizes and benchmark scores.

  2. 2

    Sort and filter

    Click any column header to sort by price, context window, or benchmark score. Use the provider filter to focus on specific vendors.

  3. 3

    Evaluate price vs. performance

    Quality and speed metrics from Artificial Analysis help you weigh cost against actual model capability.

Why Use This Pricing Comparison

  • 400+ models from 15+ providers in a single sortable table — no tab switching
  • Enriched with quality index and speed benchmarks from Artificial Analysis
  • Provider-colored badges for quick visual scanning across vendors
  • Context window and max output token data alongside pricing
  • Data sourced from official provider docs and LiteLLM open-source project

Common Use Cases

Vendor selection

Compare pricing across all major providers before committing to an API integration. Sort by cost to find the cheapest option.

Cost optimization

Find cheaper alternatives to your current model by sorting on price and checking if the quality benchmark is close enough.

Technical evaluation

Use the quality index and speed metrics to shortlist models, then compare their pricing to make a final decision.

Model comparison

Use the compare tool to see side-by-side pricing and specs for any two models.

Related Tools

Frequently Asked Questions

Common questions about LLM API pricing

点击查看文章原文
上一篇
AI价格趋势洞察 - AIscending:与AI共成长
下一篇
Agentlify Pricing
返回列表