Home Models Compare Scorecards Evals Methodology FAQ

AI Model Comparison

Compare leading AI models side-by-side with provider-verified model names and pricing references.

26 models compared Verified 2026-02-16
Pricing Performance Quick Picks Sources

Quick Reference: Pricing & Context

Input/output pricing per million tokens from provider docs. Notes are included where pricing tiers apply.

Model Pricing ($/M tokens) Context Best For
Claude Opus 4.6 (Adaptive)
Anthropic
$18.75 / $93.75
With prompt caching discounts available.
 200K
Highest intelligence tasksComplex reasoningCritical decisions
GPT-5.2 (xhigh)
OpenAI
$1.75 / $14
400K
CodingAgentsComplex tasks
Claude Opus 4.5
Anthropic
$15 / $75
200K
Complex reasoningLong-form analysisHigh-stakes drafting
GLM-5
Zhipu AI
$0.75 / $3
128K
Chinese + English workflowsValue-focused deploymentEnterprise usage
Gemini 3 Pro
Google
$2 / $12
Paid tier pricing.
 1M
Multimodal understandingAgentic tasksVibe coding
Gemini 2.5 Pro
Google
$1.25 / $10
For prompts up to 200K tokens.
 1M
Large context tasksMultimodal workflowsResearch synthesis
Claude Sonnet 4
Anthropic
$3 / $15
200K
Balanced performanceProduction workloadsGeneral-purpose tasks
DeepSeek-R1
DeepSeek
$0.55 / $2.19
128K
Budget-conscious reasoningMath-heavy tasksCost-sensitive coding
DeepSeek-V3
DeepSeek
$0.27 / $1.1
128K
Cost-effective codingGeneral tasksHigh-volume usage
Grok 4.1
xAI
$2 / $10
Via xAI API.
 128K
Real-time informationWitty responsesCurrent events
Grok 4.1 Fast
xAI
$1 / $5
Via xAI API.
 2M
Fast responsesLarge contextReal-time data
Llama 4 Scout
Meta
$0.1 / $0.3
Via API providers, pricing varies.
 10M
Extremely long contextDocument processingResearch
Llama 4 Maverick
Meta
$0.15 / $0.5
Via API providers, pricing varies.
 1M
Balanced open-sourceGeneral tasksSelf-hosting
Qwen 2.5 Max
Alibaba
$0.5 / $2
Via Alibaba Cloud.
 128K
Chinese languageMath reasoningCoding
Mistral Large 2
Mistral AI
$2 / $6
128K
European complianceMultilingual tasksEnterprise
Claude Sonnet 3.7
Anthropic
$3 / $15
200K
Extended thinkingComplex analysisCoding assistance
GPT-5 mini
OpenAI
$0.25 / $2
128K
Fast tasksHigh-volume usageCost optimization
Gemini 2.5 Flash
Google
$0.3 / $2.5
1M
Fast processingLow-latency tasksHigh throughput
Gemini 2.5 Flash-Lite
Google
$0.1 / $0.4
1M
Cost-sensitive tasksHigh-volume processingSimple queries
Claude Haiku 3.5
Anthropic
$0.8 / $4
200K
Fast responsesSimple tasksCost-conscious usage
Nova Pro
Amazon
$0.8 / $3.2
Via AWS Bedrock.
 300K
AWS integrationEnterprise workloadsMultimodal
Nova Micro
Amazon
$0.035 / $0.14
Via AWS Bedrock.
 128K
Lowest costSimple tasksHigh volume
Qwen 2.5 72B
Alibaba
$0.35 / $1.4
Via API providers.
 128K
Open-source alternativeSelf-hostingCustom fine-tuning
Mistral Small 3
Mistral AI
$0.2 / $0.6
128K
Fast processingCost-effectiveSimple tasks
Cohere Command R+
Cohere
$2.5 / $10
128K
RAG applicationsEnterprise searchTool use
Reka Core
Reka
$1 / $4
Via API providers.
 128K
Multimodal tasksVideo understandingLong context

Performance Scores

Internal task-level evaluations across coding, reasoning, and tool-use (scale: 1-10).

Model Coding Reasoning Tool-use Key Strengths
Claude Opus 4.6 (Adaptive)
Anthropic
9.6
9.7
9.4
  • Top-tier intelligence
  • Adaptive thinking
  • Best for complex reasoning
GPT-5.2 (xhigh)
OpenAI
9.6
9.5
9.5
  • Best coding performance
  • Strong agentic capabilities
  • Excellent tool integration
Claude Opus 4.5
Anthropic
9.4
9.5
9.2
  • Excellent reasoning depth
  • Strong instruction fidelity
  • Consistent long-context behavior
GLM-5
Zhipu AI
9
9.1
8.8
  • Top-tier intelligence
  • Low API cost
  • Strong bilingual support
Gemini 3 Pro
Google
9.3
9.3
9.1
  • Best multimodal model
  • Superior search integration
  • Powerful agentic capabilities
Gemini 2.5 Pro
Google
9.2
9.2
9
  • Very large context
  • Strong multimodal support
  • Competitive long-context pricing
Claude Sonnet 4
Anthropic
9.1
9
8.9
  • Great value proposition
  • Consistent quality
  • Fast response times
DeepSeek-R1
DeepSeek
8.9
9.1
8.6
  • Strong price/performance
  • Reasoning-focused behavior
  • Low token cost
DeepSeek-V3
DeepSeek
8.8
8.9
8.5
  • Excellent value
  • Good coding performance
  • Low latency
Grok 4.1
xAI
8.8
8.8
8.6
  • Real-time web access
  • Unique personality
  • Good reasoning
Grok 4.1 Fast
xAI
8.5
8.5
8.3
  • Very fast
  • Large context window
  • Real-time capabilities
Llama 4 Scout
Meta
8.6
8.7
8.4
  • Largest context window (10M)
  • Open source
  • Good for long documents
Llama 4 Maverick
Meta
8.7
8.6
8.3
  • Good performance
  • Open weights
  • Flexible deployment
Qwen 2.5 Max
Alibaba
8.8
8.8
8.5
  • Strong math skills
  • Good Chinese support
  • Competitive pricing
Mistral Large 2
Mistral AI
8.7
8.7
8.4
  • Strong multilingual
  • GDPR compliant
  • Good coding
Claude Sonnet 3.7
Anthropic
8.9
8.9
8.8
  • Strong extended thinking
  • Good coding support
  • Reliable performance
GPT-5 mini
OpenAI
8.5
8.4
8.3
  • Very low cost
  • Fast response times
  • Good for simple tasks
Gemini 2.5 Flash
Google
8.7
8.7
8.5
  • Excellent speed
  • Large context window
  • Hybrid reasoning support
Gemini 2.5 Flash-Lite
Google
8.2
8.1
7.9
  • Lowest cost option
  • Fastest output speed (499 t/s)
  • Good for simple tasks
Claude Haiku 3.5
Anthropic
8.3
8.2
8.1
  • Fast and affordable
  • Good for simple tasks
  • Reliable quality
Nova Pro
Amazon
8.5
8.5
8.3
  • AWS ecosystem
  • Good multimodal
  • Enterprise features
Nova Micro
Amazon
7.8
7.7
7.5
  • Very low cost
  • Fast inference
  • AWS integration
Qwen 2.5 72B
Alibaba
8.6
8.5
8.2
  • Open weights
  • Good performance
  • Flexible deployment
Mistral Small 3
Mistral AI
8.2
8.1
7.9
  • Very affordable
  • Fast inference
  • Good for simple tasks
Cohere Command R+
Cohere
8.3
8.4
8.6
  • Excellent RAG
  • Strong tool use
  • Enterprise focused
Reka Core
Reka
8.4
8.5
8.2
  • Strong multimodal
  • Good video understanding
  • Competitive pricing

Quick Pick Recommendations

Not sure which model? Here's our picks by use case.

💻

Best for Coding

GPT-5.2 (xhigh)

Top coding performance with excellent agentic capabilities and tool integration.

Alternative: Claude Opus 4.6 for code review quality

💰

Best on a Budget

DeepSeek-R1

Very strong reasoning with one of the lowest published API prices in this set.

Alternative: Gemini 2.5 Flash-Lite for high-volume simple tasks

📚

Best for Long Context

Llama 4 Scout

Up to 10M context window - the largest available. Ideal for massive document processing.

Alternative: Gemini 2.5 Pro for 1M context with better reasoning

🧠

Best for Reasoning

Claude Opus 4.6 (Adaptive)

Highest intelligence model with exceptional reasoning depth and adaptive thinking.

Alternative: Claude Opus 4.5 for consistent long-context behavior

🏆

Best All-Rounder

Claude Opus 4.6 (Adaptive)

Top-tier intelligence across coding, reasoning, and tool-use with adaptive thinking.

Alternative: GPT-5.2 (xhigh) for agentic tasks

🔧

Best for Tool-Use

GPT-5.2 (xhigh)

Reliable multi-step API and tooling behavior in operator-style tasks.

Alternative: Claude Sonnet 4 for balanced tool use at lower cost

Verification Sources

Official model and pricing references checked on 2026-02-16.

Model Official Sources
Claude Opus 4.6 (Adaptive)
Anthropic
Anthropic ModelsAnthropic Pricing
GPT-5.2 (xhigh)
OpenAI
OpenAI ModelsOpenAI Pricing
Claude Opus 4.5
Anthropic
Anthropic ModelsAnthropic Pricing
GLM-5
Zhipu AI
GLM-5 IntroductionZhipu Pricing
Gemini 3 Pro
Google
Gemini Model ListGemini API Pricing
Gemini 2.5 Pro
Google
Gemini Model ListGemini API Pricing
Claude Sonnet 4
Anthropic
Anthropic ModelsAnthropic Pricing
DeepSeek-R1
DeepSeek
DeepSeek Reasoning ModelDeepSeek API Pricing
DeepSeek-V3
DeepSeek
DeepSeek ModelsDeepSeek API Pricing
Grok 4.1
xAI
xAI APIGrok Documentation
Grok 4.1 Fast
xAI
xAI APIGrok Documentation
Llama 4 Scout
Meta
Meta LlamaHugging Face Llama
Llama 4 Maverick
Meta
Meta LlamaHugging Face Llama
Qwen 2.5 Max
Alibaba
Qwen ModelsAlibaba Cloud
Mistral Large 2
Mistral AI
Mistral ModelsMistral Pricing
Claude Sonnet 3.7
Anthropic
Anthropic ModelsAnthropic Pricing
GPT-5 mini
OpenAI
OpenAI ModelsOpenAI Pricing
Gemini 2.5 Flash
Google
Gemini Model ListGemini API Pricing
Gemini 2.5 Flash-Lite
Google
Gemini Model ListGemini API Pricing
Claude Haiku 3.5
Anthropic
Anthropic ModelsAnthropic Pricing
Nova Pro
Amazon
AWS BedrockNova Models
Nova Micro
Amazon
AWS BedrockNova Models
Qwen 2.5 72B
Alibaba
Qwen ModelsHugging Face Qwen
Mistral Small 3
Mistral AI
Mistral ModelsMistral Pricing
Cohere Command R+
Cohere
Cohere ModelsCohere Pricing
Reka Core
Reka
Reka AIReka Documentation

Dive Deeper into Model Performance

See detailed daily scorecards with task-level breakdowns, failure cases, and cost analysis.

View Daily Scorecards → Our Methodology