AI API pricing, compared
Compare API prices for text, TTS, image, and video models across every major provider and cloud channel — preserving each provider's native billing unit. Click any model for the full breakdown.
Headline prices today
Snapshots from our verified table below. Every row links to its official source and last-checked date.
| Model | Input /1M | Output /1M | Cached /1M | Context | Details |
|---|---|---|---|---|---|
| GPT-5.5 OpenAI | $5.00 | $30.00 | $0.50 | 1,050,000 | View → |
| Claude Opus 4.8 Anthropic | $5.00 | $25.00 | $0.50 | 1M | View → |
| Gemini 3.1 Pro Google Gemini | $2.00 | $12.00 | $0.20 | 1,048,576 | View → |
| DeepSeek V4 Pro DeepSeek | $0.435 | $0.87 | $0.0036 | 1M | View → |
| Mistral Large 3 Mistral | $0.50 | $1.50 | — | 256K | View → |
Budget picks in the same families:
-
GPT-5.4 nano
OpenAI · Classification, Low latency, High volume
$0.20 /1M in · $1.25 /1M out
-
Claude Haiku 4.5
Anthropic · Fast, Extraction, Routing
$1.00 /1M in · $5.00 /1M out
-
Gemini 3.1 Flash-Lite
Google Gemini · Low cost, High volume, Fast
$0.25 /1M in · $1.50 /1M out
Full price table (filterable)
| Model & channel | Input | Output | Cached in | Context | Best for | |
|---|---|---|---|---|---|---|
 GPT-5.5OpenAI | $5.00/1M | $30.00/1M | $0.50 | 1,050,000 | ReasoningCodingAgentic | |
| Official APIDirect | $5.00/1M | $30.00/1M | $0.50 | 1,050,000 | ||
 Azure OpenAICloud | $5.00/1M | $30.00/1M | $0.50 | 1,050,000 | ||
 OpenRouterRouter | $5.00/1M | $30.00/1M | — | 1,050,000 | ||
| GPT-5.5 ProOpenAI | $30.00/1M | $180.00/1M | — | 1,050,000 | ReasoningHigh-stakes | |
| Official APIDirect | $30.00/1M | $180.00/1M | — | 1,050,000 | ||
| OpenRouterRouter | $30.00/1M | $180.00/1M | — | 1,050,000 | ||
| GPT-5.4OpenAI | $2.50/1M | $15.00/1M | $0.25 | 1,050,000 | ReasoningCodingGeneral | |
| Official APIDirect | $2.50/1M | $15.00/1M | $0.25 | 1,050,000 | ||
| Azure OpenAICloud | $2.50/1M | $15.00/1M | $0.25 | 1,050,000 | ||
| OpenRouterRouter | $2.50/1M | $15.00/1M | — | 1,050,000 | ||
| GPT-5.4 miniOpenAI | $0.75/1M | $4.50/1M | $0.075 | 400,000 | Low latencyCodingAgents at scale | |
| Official APIDirect | $0.75/1M | $4.50/1M | $0.075 | 400,000 | ||
| Azure OpenAICloud | $0.75/1M | $4.50/1M | $0.075 | 400,000 | ||
| OpenRouterRouter | $0.75/1M | $4.50/1M | — | 400,000 | ||
| GPT-5.4 nanoOpenAI | $0.20/1M | $1.25/1M | $0.02 | 400,000 | ClassificationLow latencyHigh volume | |
| Official APIDirect | $0.20/1M | $1.25/1M | $0.02 | 400,000 | ||
| Azure OpenAICloud | $0.20/1M | $1.25/1M | $0.02 | 400,000 | ||
| OpenRouterRouter | $0.20/1M | $1.25/1M | — | 400,000 | ||
| GPT-5.3-CodexOpenAI | $1.75/1M | $14.00/1M | $0.175 | 400,000 | Agentic codingSoftware engineeringTool use | |
| Official APIDirect | $1.75/1M | $14.00/1M | $0.175 | 400,000 | ||
| Azure OpenAICloud | $1.75/1M | $14.00/1M | $0.175 | 400,000 | ||
| OpenRouterRouter | $1.75/1M | $14.00/1M | — | 400,000 | ||
 Claude Opus 4.8Anthropic | $5.00/1M | $25.00/1M | $0.50 | 1M | Hard reasoningAgentsCoding | |
| Claude APIDirect | $5.00/1M | $25.00/1M | $0.50 | 1M | ||
 AWS BedrockCloud | $5.00/1M | $25.00/1M | $0.50 | 1M | ||
 Google Vertex AICloud | $5.00/1M | $25.00/1M | $0.50 | 1M | ||
| OpenRouterRouter | $5.00/1M | $25.00/1M | — | 1M | ||
| Claude Opus 4.7Anthropic | $5.00/1M | $25.00/1M | $0.50 | 1M | Hard reasoningAgents | |
| Claude APIDirect | $5.00/1M | $25.00/1M | $0.50 | 1M | ||
| AWS BedrockCloud | $5.00/1M | $25.00/1M | $0.50 | 1M | ||
| Google Vertex AICloud | $5.00/1M | $25.00/1M | $0.50 | 1M | ||
| OpenRouterRouter | $5.00/1M | $25.00/1M | — | 1M | ||
| Claude Sonnet 4.6Anthropic | $3.00/1M | $15.00/1M | $0.30 | 1M | CodingAgentsProduction | |
| Claude APIDirect | $3.00/1M | $15.00/1M | $0.30 | 1M | ||
| Google Vertex AICloud | $3.00/1M | $15.00/1M | $0.30 | 1M | ||
| OpenRouterRouter | $3.00/1M | $15.00/1M | — | 1M | ||
| Claude Sonnet 4.5Anthropic | $3.00/1M | $15.00/1M | $0.30 | 1M | CodingAgents | |
| Claude APIDirect | $3.00/1M | $15.00/1M | $0.30 | 1M | ||
| Google Vertex AICloud | $3.00/1M | $15.00/1M | $0.30 | 1M | ||
| OpenRouterRouter | $3.00/1M | $15.00/1M | — | 1M | ||
| Claude Haiku 4.5Anthropic | $1.00/1M | $5.00/1M | $0.10 | 200K | FastExtractionRouting | |
| Claude APIDirect | $1.00/1M | $5.00/1M | $0.10 | 200K | ||
| AWS BedrockCloud | $1.00/1M | $5.00/1M | $0.10 | 200K | ||
| Google Vertex AICloud | $1.00/1M | $5.00/1M | $0.10 | 200K | ||
| OpenRouterRouter | $1.00/1M | $5.00/1M | — | 200K | ||
 Gemini 3.1 ProGoogle Gemini | $2.00/1M | $12.00/1M | $0.20 | 1,048,576 | ReasoningAgenticLong context | |
| Gemini Developer APIDirect | $2.00/1M | $12.00/1M | $0.20 | 1,048,576 | ||
| Google Vertex AICloud | $2.00/1M | $12.00/1M | $0.20 | 1,048,576 | ||
| OpenRouterRouter | $2.00/1M | $12.00/1M | — | 1,048,576 | ||
| Gemini 3.5 FlashGoogle Gemini | $1.50/1M | $9.00/1M | $0.15 | 1,048,576 | CodingAgenticFast | |
| Gemini Developer APIDirect | $1.50/1M | $9.00/1M | $0.15 | 1,048,576 | ||
| Google Vertex AICloud | $1.50/1M | $9.00/1M | $0.15 | 1,048,576 | ||
| OpenRouterRouter | $1.50/1M | $9.00/1M | — | 1,048,576 | ||
| Gemini 3.1 Flash-LiteGoogle Gemini | $0.25/1M | $1.50/1M | $0.025 | 1,048,576 | Low costHigh volumeFast | |
| Gemini Developer APIDirect | $0.25/1M | $1.50/1M | $0.025 | 1,048,576 | ||
| Google Vertex AICloud | $0.25/1M | $1.50/1M | $0.025 | 1,048,576 | ||
| OpenRouterRouter | $0.25/1M | $1.50/1M | — | 1,048,576 | ||
| Gemini 2.5 ProGoogle Gemini | $1.25/1M | $10.00/1M | $0.125 | 1,048,576 | ReasoningLong contextMultimodal | |
| Gemini Developer APIDirect | $1.25/1M | $10.00/1M | $0.125 | 1,048,576 | ||
| Google Vertex AICloud | $1.25/1M | $10.00/1M | $0.13 | 1,048,576 | ||
| OpenRouterRouter | $1.25/1M | $10.00/1M | — | 1,048,576 | ||
| Gemini 2.5 FlashGoogle Gemini | $0.30/1M | $2.50/1M | $0.03 | 1,048,576 | Low latencyHigh volumeMultimodal | |
| Gemini Developer APIDirect | $0.30/1M | $2.50/1M | $0.03 | 1,048,576 | ||
| Google Vertex AICloud | $0.30/1M | $2.50/1M | $0.03 | 1,048,576 | ||
| OpenRouterRouter | $0.30/1M | $2.50/1M | — | 1,048,576 | ||
| Gemini 2.5 Flash-LiteGoogle Gemini | $0.10/1M | $0.40/1M | $0.01 | 1,048,576 | Lowest costHigh volume | |
| Gemini Developer APIDirect | $0.10/1M | $0.40/1M | $0.01 | 1,048,576 | ||
| Google Vertex AICloud | $0.10/1M | $0.40/1M | $0.01 | 1,048,576 | ||
| OpenRouterRouter | $0.10/1M | $0.40/1M | — | 1,048,576 | ||
 Grok 4.3xAI | $1.25/1M | $2.50/1M | $0.20 | 1M | ReasoningTool callingAgents | |
| Official APIDirect | $1.25/1M | $2.50/1M | $0.20 | 1M | ||
| OpenRouterRouter | $1.25/1M | $2.50/1M | — | 1M | ||
| Azure AI FoundryCloud | $1.25/1M | $2.50/1M | — | 200K (Foundry cap) | ||
 DeepSeek V4 FlashOpenDeepSeek | $0.14/1M | $0.28/1M | $0.0028 | 1M | Open weightLow costHigh volume | |
| Official APIDirect | $0.14/1M | $0.28/1M | $0.0028 | 1M | ||
| OpenRouterRouter | $0.09/1M | $0.18/1M | — | 1M | ||
| DeepSeek V4 ProOpenDeepSeek | $0.435/1M | $0.87/1M | $0.0036 | 1M | Open weightReasoningCoding | |
| Official APIDirect | $0.435/1M | $0.87/1M | $0.0036 | 1M | ||
| OpenRouterRouter | $0.435/1M | $0.87/1M | — | 1M | ||
 Together AIHost | $2.10/1M | $4.40/1M | $0.20 | 512K | ||
 FireworksHost | $1.74/1M | $3.48/1M | — | 1M | ||
 Mistral Large 3OpenMistral | $0.50/1M | $1.50/1M | — | 256K | Open weightAgenticMultilingual | |
| La PlateformeDirect | $0.50/1M | $1.50/1M | — | 256K | ||
| AWS BedrockCloud | $0.50/1M | $1.50/1M | — | 128K | ||
| Mistral Medium 3.5Mistral · La Plateforme | $1.50/1M | $7.50/1M | — | 128K | BalancedEnterprise | |
| Mistral Small 4OpenMistral · La Plateforme | $0.10/1M | $0.30/1M | — | 128K | Open weightLow costEdge | |
| Devstral 2OpenMistral · La Plateforme | $0.40/1M | $2.00/1M | — | 256K | Open weightCodingAgentic | |
 Llama 3.3 70BOpenMeta | $0.10/1M | $0.32/1M | — | 131K | Open weightGeneralCheap | |
| OpenRouterRouter | $0.10/1M | $0.32/1M | — | 131K | ||
 GroqHost | $0.59/1M | $0.79/1M | — | 128K | ||
| AWS BedrockCloud | $0.72/1M | $0.72/1M | — | 128K | ||
| FireworksHost | $0.90/1M | $0.90/1M | $0.45 | 131K | ||
| Together AIHost | $1.04/1M | $1.04/1M | — | 131K | ||
| Llama 4 ScoutOpenMeta | $0.10/1M | $0.30/1M | — | 10M | Open weightUltra-long contextMultimodal | |
| OpenRouterRouter | $0.10/1M | $0.30/1M | — | 10M | ||
| GroqHost | $0.11/1M | $0.34/1M | — | 128K | ||
| FireworksHost | $0.15/1M | $0.60/1M | $0.075 | 131K | ||
| Llama 4 MaverickOpenMeta | $0.15/1M | $0.60/1M | — | 1M | Open weightReasoningMultimodal | |
| OpenRouterRouter | $0.15/1M | $0.60/1M | — | 1M | ||
| FireworksHost | $0.22/1M | $0.88/1M | $0.11 | 131K | ||
| AWS BedrockCloud | $0.24/1M | $0.97/1M | — | 1M | ||
| Together AIHost | $0.27/1M | $0.85/1M | — | 1M | ||
 Qwen3 235B A22BOpenAlibaba | $0.09/1M | $0.10/1M | — | 262K | Open weightReasoningMoE | |
| OpenRouterRouter | $0.09/1M | $0.10/1M | — | 262K | ||
| Together AIHost | $0.20/1M | $0.60/1M | — | 131K | ||
| FireworksHost | $0.22/1M | $0.88/1M | $0.11 | 131K | ||
| Qwen3 Coder 480BOpenAlibaba · OpenRouter | $0.22/1M | $1.80/1M | — | 1M | Open weightCodingAgentic | |
| Qwen3 32BOpenAlibaba · Groq | $0.29/1M | $0.59/1M | — | 131K | Open weightFastDense | |
 Gemma 3 27BOpenGoogle · OpenRouter | $0.08/1M | $0.16/1M | — | 131K | Open weightSingle-GPUMultimodal | |
| Gemma 3 12BOpenGoogle · OpenRouter | $0.05/1M | $0.15/1M | — | 131K | Open weightCheapSmall | |
 Phi-4 (14B)OpenMicrosoft · OpenRouter | $0.065/1M | $0.14/1M | — | 16K | Open weightSmallEfficient | |
 Kimi K2.6OpenMoonshot | $0.95/1M | $4.00/1M | $0.16 | 262K | Open weightReasoningMultimodal | |
| Official APIDirect | $0.95/1M | $4.00/1M | $0.16 | 262K | ||
| OpenRouterRouter | $0.68/1M | $3.41/1M | — | 262K | ||
| Together AIHost | $1.20/1M | $4.50/1M | $0.20 | 262K | ||
| FireworksHost | $0.95/1M | $4.00/1M | — | 262K | ||
| Kimi K2.7 CodeOpenMoonshot | $0.95/1M | $4.00/1M | $0.19 | 262K | Open weightCodingAgentic+1 | |
| OpenRouterRouter | $0.74/1M | $3.50/1M | — | 262K | ||
| Official APIDirect | $0.95/1M | $4.00/1M | $0.19 | 262K | ||
| —/1M | —/1M | — | 262K | |||
 MiniMax M3OpenMiniMax | $0.30/1M | $1.20/1M | $0.06 | 1,048,576 | Open weightAgenticCoding+1 | |
| Official APIDirect | $0.30/1M | $1.20/1M | $0.06 | 1,048,576 | ||
| $0.60/1M | $2.40/1M | $0.12 | 1,048,576 | |||
| OpenRouterRouter | $0.30/1M | $1.20/1M | — | 1,048,576 | ||
| MiniMax M2.7OpenMiniMax | $0.30/1M | $1.20/1M | $0.06 | Needs review | Open weightAgenticAffordable | |
| Official APIDirect | $0.30/1M | $1.20/1M | $0.06 | Needs review | ||
| $0.60/1M | $2.40/1M | $0.06 | Needs review | |||
| OpenRouterRouter | $0.30/1M | $1.20/1M | — | Needs review | ||
| MiniMax M2.5OpenMiniMax | $0.30/1M | $1.20/1M | $0.03 | Needs review | Open weightLegacyAffordable | |
| Official APIDirect | $0.30/1M | $1.20/1M | $0.03 | Needs review | ||
| $0.60/1M | $2.40/1M | $0.03 | Needs review | |||
 GLM-5.2OpenZhipu | $1.40/1M | $4.40/1M | $0.26 | 1,048,576 | Open weightCodingReasoning+1 | |
| Official API (Z.ai)Direct | $1.40/1M | $4.40/1M | $0.26 | 1,048,576 | ||
| $1.40/1M | $4.40/1M | $0.25 | 1,048,576 | |||
| OpenRouterRouter | $1.40/1M | $4.40/1M | — | 1,048,576 | ||
| GLM-5.1OpenZhipu | $1.40/1M | $4.40/1M | $0.26 | Needs review | Open weightCodingAgentic | |
| OpenRouterRouter | $0.98/1M | $3.08/1M | — | Needs review | ||
| Official API (Z.ai)Direct | $1.40/1M | $4.40/1M | $0.26 | Needs review | ||
| Qwen3.7 MaxOpenAlibaba | $1.25/1M | $3.75/1M | $0.13 | Needs review | Open weightReasoningAgentic | |
| Together AIHost | $1.25/1M | $3.75/1M | $0.13 | Needs review | ||
| OpenRouterRouter | $1.25/1M | $3.75/1M | — | Needs review | ||
| Qwen3.7 PlusAlibaba · Requesty | $0.32/1M | $1.28/1M | — | Needs review | AffordableGeneralAlibaba |
Frequently asked questions
What is the cheapest AI API in 2026?
As of 2026-06-18, the cheapest verified text API is Gemini 2.5 Flash-Lite at $0.10 per 1M input tokens and $0.40 per 1M output tokens, with cached input at $0.01 per 1M; DeepSeek V4 Flash is cheaper on output at $0.14/$0.28, and $0.0028/1M for cache hits. For image generation, MiniMax image-01 is the lowest official row we track at $0.0035/image; OpenAI's cheapest image tier is GPT Image 1 Mini from about $0.005/image. For video, Google's Veo 3.1 Lite is the lowest official row we verified at $0.05/sec for 720p.
How much does GPT-5 cost per 1M tokens?
As of 2026-06-17, GPT-5.5 (OpenAI direct) costs $5.00 per 1M input tokens and $30.00 per 1M output tokens, with cached input at $0.50 per 1M. GPT-5.4 is $2.50/$15.00. GPT-5.4 mini is $0.75/$4.50. GPT-5.4 nano is $0.20/$1.25. Prices are identical on Azure OpenAI for the same models.
How much does Claude API cost per 1M tokens?
As of 2026-06-17, Claude Opus 4.8 costs $5.00 per 1M input tokens and $25.00 per 1M output tokens (cached read $0.50). Claude Sonnet 4.6 is $3.00/$15.00 (cached read $0.30). Claude Haiku 4.5 is $1.00/$5.00 (cached read $0.10).
What is prompt caching and how much does it save?
Prompt caching reuses a previously seen input prefix at a discounted rate. On GPT-5.5 cached input is $0.50 per 1M versus $5.00 fresh — a 90% discount. On Claude Sonnet 4.6 cached read is $0.30 versus $3.00 fresh — also 90% off. Anthropic additionally separates 5-minute and 1-hour cache writes.
Is OpenRouter the same price as the official API?
Often yes. As of 2026-06-17, OpenRouter lists Claude Sonnet 4.6 at $3.00/$15.00 per 1M — identical to Anthropic direct. The trade-off: OpenRouter hides fine-grained cache read/write tiers, so if your workload depends on cache economics, the direct API exposes the cleaner price.
Which AI coding tool costs $20/month?
As of 2026-06-17, Cursor Pro is $20/month, ChatGPT Plus is $20/month, Claude Pro is $20/month (with confirmed $200/year annual), and Perplexity Pro is $20/month. GitHub Copilot Pro is the cheapest at $10/user/month.