LLM pricing dataset

AI model pricing table

Browse 281 crawler-readable model rows across 38 providers. Prices are USD per 1,000,000 tokens.

Models281
Providers38
Last updated2026-06-21
ModelProviderInput $/MOutput $/MContextTierBest for
Hunyuan-LiteTencent$0$0256,000BudgetFree Tencent Cloud tier for high-volume chat and classification
Owl AlphaOpenRouter$0$01,048,756FlagshipFree alpha — high-performance agentic foundation model with native tool use and ~1.05M context. Prompts/completions may be logged.
Gemma 2 9BGoogle$0.030$0.0608,192Open SourceSmaller open-weight Gemma 2 for edge and on-device
Nova MicroAmazon$0.035$0.140128,000BudgetCheapest text-only Bedrock model for high-volume workloads
Command R7BCohere$0.037$0.150128,000BudgetUltra-cheap, low-latency high-volume tasks
Ministral 3BMistral$0.040$0.040128,000BudgetUltra-cheap 3B model for classification and routing
Phi-4 MiniMicrosoft$0.040$0.100128,000BudgetCompact 3.8B Phi-4 sibling for low-latency reasoning
Llama 3 8BMeta$0.050$0.0808,192Open SourcePrevious-gen 8B open-weight for chat and edge
Llama 3.1 8BMeta$0.050$0.080128,000Open SourceOpen-weight 8B for low-latency chat and edge
Llama 3.1 8B Instant (Groq)Groq$0.050$0.080128,000BudgetUltra-fast LPU-hosted 8B for sub-second responses
Gemma 3 12BGoogle$0.050$0.100128,000Open SourceMid-size open-weight Gemma 3 for self-hosted chat and reasoning
Qwen3 TurboQwen$0.050$0.200131,072BudgetHigh-throughput cost-efficient inference
GLM-4.5 FlashZhipu$0.050$0.200128,000BudgetHigh-volume classification and retrieval
GPT-5.4 NanoOpenAI$0.050$0.400128,000BudgetClassification, routing, simple tasks
Qwen3 FlashQwen$0.050$0.400128,000BudgetUltra-cheap classification, chat, and high-volume tasks
GPT-5 NanoOpenAI$0.050$0.400400,000BudgetFast, low-cost OpenAI model for simple or high-volume tasks
Nova CanvasAmazon$0.060$0.0600Mid-tierImage generation and editing on Amazon Bedrock
Hunyuan-StandardTencent$0.060$0.13032,000BudgetDefault Tencent Cloud model for everyday chat and RAG
Hy3 LiteTencent$0.060$0.180128,000BudgetUltra-fast MoE for chat, classification, and high-volume apps
MiMo LiteXiaomi$0.060$0.180128,000BudgetUltra-affordable entry model for high-volume apps
Nova LiteAmazon$0.060$0.240300,000BudgetAffordable multimodal Bedrock model — text, vision, video understanding
Phi-4Microsoft$0.065$0.14016,000BudgetLow-latency coding, reasoning, and chat on Azure AI Foundry
Grok ImagexAI$0.070$0.0700Open SourceAurora image generation — photorealistic text-to-image
abab6.5t-chatMiniMax$0.070$0.0708,192BudgetTiny abab6.5 variant for ultra-low-latency apps
Mistral 7B InstructMistral$0.070$0.07032,000Open SourceClassic open-weight 7B instruct model
Gemma 2 27BGoogle$0.070$0.1408,192Open SourceOpen-weight 27B model for self-hosted general purpose
Gemini 2.0 Flash-LiteGoogle$0.075$0.3001,000,000BudgetUltra-affordable high-volume tasks
Gemini 1.5 FlashGoogle$0.075$0.3001,000,000BudgetFast, affordable previous-gen model with 1M context
Nemotron Mini 4B InstructNvidia$0.080$0.0804,096BudgetTiny on-device Nemotron for edge inference
Gemma 3 27BGoogle$0.080$0.160128,000Open SourceOpen-weight multimodal model for text and vision tasks
MiniMax Text-LiteMiniMax$0.080$0.240128,000BudgetUltra-affordable entry model for chat and high-volume apps
Llama 4 Scout (Fireworks AI)Fireworks AI$0.090$0.2501,000,000BudgetFireworks-hosted Llama 4 Scout — ultra-fast inference
Ministral 8BMistral$0.100$0.100128,000BudgetEdge-class 8B model — low-latency on-device chat
Mistral Embed 2Mistral$0.100$0.10032,000BudgetMultilingual embeddings for RAG and clustering
Mistral ModerationMistral$0.100$0.10032,000BudgetContent moderation across 9 safety categories
Granite 3.3 2B InstructIBM$0.100$0.100128,000BudgetUltra-cheap on-device/edge inference, classification, summarization on watsonx.ai
Embed v3 EnglishCohere$0.100$0.100512Budget1024-dim English embeddings for semantic search and RAG
Granite Vision 3.2 2BIBM$0.100$0.10016,000BudgetCompact vision-language model for document and chart understanding
Yi-Coder 9B01.AI$0.100$0.100128,000Open SourceOpen-weight 9B code model — competitive with 33B on benchmarks
Llama 3.1 8B (Cerebras)Cerebras$0.100$0.100128,000BudgetWafer-scale 8B for ultra-low-latency chat
Llama 4 ScoutMeta$0.100$0.25010,000,000Open SourceFree self-host, efficient
Grok 4.1 MinixAI$0.100$0.3001,000,000BudgetSmallest 4.1 variant — RAG and high-volume classification
Devstral SmallMistral$0.100$0.300128,000BudgetOpen-weight agentic coder for IDEs and local dev
Nemotron Nano 9B v2Nvidia$0.100$0.300128,000BudgetCheapest fast Nemotron — high-throughput agents and RAG
MiMo V2.5 LiteXiaomi$0.100$0.300128,000BudgetFast, low-cost variant for chat, classification, and routing
Llama 4 Scout (Together AI)Together AI$0.100$0.3001,000,000BudgetHosted open-weight inference — fast, affordable general purpose
Phi-5Microsoft$0.100$0.300256,000Mid-tierNext-gen Phi small-model flagship — strong quality per dollar
GPT-4.1 NanoOpenAI$0.100$0.4001,000,000BudgetHigh-volume, low-latency workloads
Gemini 2.5 Flash-LiteGoogle$0.100$0.4001,000,000BudgetCheapest major provider option
Gemini 2.0 FlashGoogle$0.100$0.4001,000,000BudgetPrevious-gen fast model
Hunyuan-TurboSTencent$0.110$0.420256,000Mid-tierHybrid Transformer-Mamba balance of speed and quality
Nano (STT)AssemblyAI$0.120$00BudgetBudget STT: $0.12 per audio hour
Embed v4 MultilingualCohere$0.120$0.1204,096Budget1024-dim multilingual embeddings across 100+ languages
Mistral Small 3.1 (Fireworks AI)Fireworks AI$0.130$0.500128,000BudgetFireworks-hosted Mistral Small 3.1 for high-volume agents
Yi-Lightning01.AI$0.140$0.14016,384FlagshipLow-latency flagship for chat and reasoning
abab6.5s-chatMiniMax$0.140$0.140245,000BudgetLighter, faster abab6.5 sibling for high-volume chat
DeepSeek V2.5DeepSeek$0.140$0.280128,000Mid-tierPrevious-gen DeepSeek for general chat and coding
DeepSeek Coder V2DeepSeek$0.140$0.280128,000Mid-tierCode-specialized MoE for IDE completion and repo-scale tasks
DeepSeek V4 FlashDeepSeek$0.140$0.550128,000Budget284B MoE (13B active), high-throughput chat/RAG/extraction
Hunyuan-T1Tencent$0.140$0.550128,000ReasoningHybrid Mamba-Transformer reasoning model with strong math and coding
Codestral EmbedMistral$0.150$0.1508,192BudgetCode embeddings for retrieval and semantic search
Pixtral 12BMistral$0.150$0.150128,000Open SourceOpen-weight 12B vision-language model
Mistral NeMoMistral$0.150$0.150128,000Open SourceOpen-weight 12B model co-developed with NVIDIA
Mistral Small 3.1Mistral$0.150$0.600128,000BudgetHigh-throughput budget tasks
Command RCohere$0.150$0.600128,000Mid-tierProduction RAG at budget pricing
Kimi K2 MiniMoonshot$0.150$0.600128,000BudgetDistilled K2 for low-latency, high-throughput chat
GPT-4o Mini (GitHub Models)GitHub Models$0.150$0.600128,000BudgetAffordable Azure-hosted small multimodal model
QwQ 32BQwen$0.150$0.600131,072ReasoningOpen-weight reasoning model with extended chain-of-thought
Phi-3.5 MoEMicrosoft$0.160$0.640131,072Mid-tier16x3.8B sparse MoE — strong quality-per-dollar on Azure AI Foundry
Qwen 2.5 Coder 32BQwen$0.180$0.180131,072Open SourceCode-specialized 32B for completion and refactoring
Llama 4 Maverick (Fireworks AI)Fireworks AI$0.180$0.5001,000,000Mid-tierFireworks-hosted Llama 4 Maverick — low-latency reasoning
Qwen 2.5 32B InstructQwen$0.200$0.200131,072Open SourceMid-size Qwen 2.5 for production agents
Sonar Small OnlinePerplexity$0.200$0.200127,000BudgetCheapest grounded search model with citations
Granite Code 8B InstructIBM$0.200$0.200128,000BudgetCode-specialized 8B for enterprise IDE integrations on watsonx.ai
Yi-2 Lightning01.AI$0.200$0.20032,768FlagshipNext-gen Yi — low-latency bilingual flagship for chat and reasoning
Granite 4 8B InstructIBM$0.200$0.200256,000Open SourceHybrid Mamba-Transformer enterprise instruct on watsonx.ai
Jamba 1.7 MiniAI21$0.200$0.400256,000BudgetCost-efficient hybrid SSM for long-context RAG
Grok 4.1 FastxAI$0.200$0.5002,000,000BudgetUltra-cheap with massive 2M context
Llama 4 MaverickMeta$0.200$0.6001,000,000Open SourceFree self-host, strong general purpose
Mistral SabaMistral$0.200$0.60032,000Mid-tierArabic + South-Asian language specialist
Llama 4 Maverick (Together AI)Together AI$0.200$0.6001,000,000Mid-tierHosted open-weight inference — strong reasoning and coding
Llama 3.3 70BMeta$0.200$0.600128,000Open Source70B instruction-tuned open-weight — strong general purpose
Llama 3.1 70BMeta$0.200$0.600128,000Open SourceOpen-weight 70B for production agents and reasoning
DeepSeek Coder V3DeepSeek$0.200$0.800256,000FlagshipCode-specialized MoE for repo-scale completion and refactoring
GLM-4.5 AirZhipu$0.200$1.10128,000Mid-tierLightweight MoE for cost-sensitive agentic workloads
Grok Code Fast 1xAI$0.200$1.50256,000Mid-tierCoding-specialized Grok for IDEs and agentic dev
Hunyuan-Standard-256KTencent$0.210$0.840256,000Mid-tierLong-context variant of Hunyuan Standard for document QA
Llama 3.3 70B (DeepInfra)DeepInfra$0.230$0.400128,000Mid-tierDeepInfra-hosted Llama 3.3 70B at aggressive per-token pricing
Mixtral 8x7BMistral$0.240$0.24032,000Open SourceSparse MoE 8x7B — efficient open-weight workhorse
Mixtral 8x7B (Groq)Groq$0.240$0.24032,000Mid-tierLPU-hosted Mixtral 8x7B with extreme throughput
Qwen3 32B (Fireworks AI)Fireworks AI$0.250$0.700128,000Mid-tierFireworks-hosted Qwen3 32B optimized for throughput
Llama 4.1 70BMeta$0.250$0.7001,000,000Open SourceRefreshed 70B with 1M context — strong general purpose
MiMo TextXiaomi$0.250$0.750256,000Mid-tierMid-tier reasoning, coding, and long-context RAG
Claude 3 HaikuAnthropic$0.250$1.25200,000BudgetCheapest Claude 3 — high-volume chat and extraction
Gemini 3.1 Flash-LiteGoogle$0.250$1.501,000,000BudgetBudget with newer capabilities
GPT-5 MiniOpenAI$0.250$2.00400,000Mid-tierLower-cost GPT-5 for strong general performance
DeepSeek V3DeepSeek$0.270$1.10128,000Flagship671B MoE flagship — strong coding and reasoning at open-weight prices
DeepSeek V3.2DeepSeek$0.280$0.420128,000BudgetBest value overall, 90% cache discount
Qwen3 32B (Groq)Groq$0.290$0.590128,000Mid-tierLPU-accelerated Qwen3 32B — extreme throughput for agents
Speechmatics StandardSpeechmatics$0.300$00BudgetSTT: $0.30 per audio hour — 50+ languages with diarization
abab6.5-chatMiniMax$0.300$0.300245,000Mid-tierTrillion-parameter MoE chat model with long context
Llama 3 70BMeta$0.300$0.4008,192Open SourcePrevious-gen Llama 3 70B instruct
Grok 3 MinixAI$0.300$0.500131,072BudgetBudget Grok 3 for chat and lightweight agents
Nemotron 3 Super 120BNvidia$0.300$0.800262,000ReasoningHybrid Mamba-Transformer MoE — agentic reasoning, long context
Qwen3 32B (Together AI)Together AI$0.300$0.800128,000Mid-tierHosted Qwen3 32B for production agents and tool use
Grok 4 MinixAI$0.300$0.900256,000BudgetCheaper Grok 4 distilled for high-throughput chat and tools
Codestral 25.10Mistral$0.300$0.900256,000Mid-tierCode generation, completion, and FIM across 80+ languages
Sarvam-1Sarvam AI$0.300$0.9008,192BudgetCompact Indic foundation model hosted in India
Hunyuan ProTencent$0.300$1.0032,000FlagshipProduction Hunyuan flagship for everyday agents and RAG
MiniMax M2MiniMax$0.300$1.201,000,000Flagship230B MoE optimized for agentic coding and end-to-end tool use
Gemini 2.5 FlashGoogle$0.300$2.501,000,000BudgetHigh-volume, budget apps
Qwen 2.5 72B InstructQwen$0.350$0.400131,072Open SourceOpen-weight 72B instruct — strong multilingual reasoning
Codestral 26Mistral$0.350$1.00256,000Mid-tierRefreshed Codestral for completion, FIM, and 80+ languages
MiniMax Text-01MiniMax$0.350$1.05256,000FlagshipLong-context text flagship — strong reasoning and RAG
Hy3 PreviewTencent$0.350$1.05128,000FlagshipFrontier preview — strong reasoning, coding, and agentic workflows
MiMo AudioXiaomi$0.350$1.0532,000FlagshipEnd-to-end speech + text model for voice agents and audio reasoning
Universal-2AssemblyAI$0.370$00FlagshipSpeech-to-text: $0.37 per audio hour
Llama 3.3 Nemotron Super 49BNvidia$0.400$0.400128,000FlagshipNVIDIA-tuned Llama for tool use and instruction following
Llama 3.1 Nemotron 70B InstructNvidia$0.400$0.400128,000FlagshipNVIDIA-tuned Llama 70B with strong arena scores
abab7-chatMiniMax$0.400$0.400256,000Mid-tierNext-gen abab chat model with extended context and stronger reasoning
Qwen3 32B (Cerebras)Cerebras$0.400$0.800128,000Mid-tierWafer-scale Qwen3 32B at 2000+ tokens/sec inference
Reka EdgeReka$0.400$1.0064,000BudgetOn-device / edge-deployable multimodal model
Qwen3 PlusQwen$0.400$1.20256,000Mid-tierBalanced quality and price for production apps
Hunyuan-TurboTencent$0.400$1.20256,000FlagshipProduction MoE flagship — fast, multilingual, function calling
MiMo VisionXiaomi$0.400$1.20128,000FlagshipVision-language MoE for multimodal reasoning and image understanding
Grok 5 MinixAI$0.400$1.501,000,000Mid-tierCheaper distilled Grok 5 for high-throughput chat and tools
GPT-5.4 MiniOpenAI$0.400$1.60128,000Mid-tierProduction apps, moderate complexity
GPT-4.1 MiniOpenAI$0.400$1.601,000,000Mid-tierStructured tasks, cost-efficient coding
Mistral Medium 3Mistral$0.400$2.00131,000Mid-tierBalanced workloads, function calling
Devstral MediumMistral$0.400$2.00256,000Mid-tierAgentic coding — long-context repo edits and tool use
Kimi K1.5Moonshot$0.400$2.00128,000Mid-tierLong-context multimodal reasoning with RL post-training
MiniMax M1MiniMax$0.400$2.201,000,000Reasoning456B MoE reasoning model with 1M-token context — first open M-series
MiniMax VL-01MiniMax$0.450$1.35128,000FlagshipVision-language MoE for multimodal reasoning and image understanding
MiMo V2.5Xiaomi$0.450$1.35256,000FlagshipOpen-weight MoE flagship — strong reasoning, coding, and RAG
Granite 3.3 8B InstructIBM$0.500$0.500128,000Open SourceEnterprise RAG, summarization, function calling on watsonx.ai
Yi-Medium01.AI$0.500$0.50016,384Mid-tierBalanced cost-quality model
Yi-VL Plus01.AI$0.500$0.50016,384Mid-tierBilingual vision-language model for charts and OCR
Jamba InstructAI21$0.500$0.700256,000Mid-tierInstruction-tuned variant for enterprise deployments
Magistral SmallMistral$0.500$1.50128,000Mid-tierOpen-weight 24B reasoning model — math, science, code
Sarvam-MSarvam AI$0.500$1.5032,768Mid-tierIndic-first 24B reasoning model across 10+ Indian languages
Hunyuan-LargeTencent$0.500$1.50256,000Open Source389B open-weight MoE — 52B active params, strong general reasoning
Hunyuan-VisionTencent$0.500$1.5032,000FlagshipMultimodal vision-language model for image + text reasoning
MiMo CoderXiaomi$0.500$1.50128,000FlagshipCode-specialized MiMo for repo-scale coding and refactoring
Aya Expanse 32BCohere$0.500$1.50128,000Mid-tier23-language open-weight model from Cohere For AI
Qwen3 VisionQwen$0.500$1.50256,000FlagshipVision-language Qwen3 for charts, OCR, and UI understanding
Mistral Large 3Mistral$0.500$2.00131,000FlagshipEuropean provider, multilingual
GLM-Z1Zhipu$0.500$2.00128,000ReasoningReasoning-tuned GLM with extended chain-of-thought
MiniMax M3MiniMax$0.500$2.204,000,000FlagshipFlagship MoE with 4M-token context — agentic reasoning and tool use
Kimi LatestMoonshot$0.500$2.50200,000Mid-tierAlways-updated production endpoint tracking the newest Kimi
Gemini 3 FlashGoogle$0.500$3.001,000,000Mid-tierFast reasoning, good value
Hunyuan-CodeTencent$0.550$1.65128,000FlagshipCode-specialized Hunyuan for repo-scale coding and refactoring
Hunyuan-RoleTencent$0.550$1.6532,000Mid-tierRole-play tuned variant for character agents and companions
Hunyuan-FunctionCallTencent$0.550$1.6532,000Mid-tierTool-use tuned variant optimized for function calling pipelines
DeepSeek R1DeepSeek$0.550$2.19128,000ReasoningReasoning at budget prices
DeepSeek V4 ProDeepSeek$0.550$2.20256,000Flagship1.6T MoE flagship (49B active), frontier reasoning at open-weight prices
Kimi K2 InstructMoonshot$0.550$2.20262,144FlagshipOpen-weight 1T MoE instruct model — coding, tool use, agents
MiMo ReasonerXiaomi$0.550$2.20128,000ReasoningReasoning-tuned MoE for math, science, and multi-step problems
Llama 3.3 70B (Groq)Groq$0.590$0.790128,000Mid-tierLPU-accelerated Llama 3.3 70B — among the fastest tokens/sec available
Llama 3.3 70B (SambaNova)SambaNova$0.600$1.20128,000Mid-tierRDU-accelerated Llama 3.3 70B for fast enterprise inference
Pharia-1-LLM-7B-ControlAleph Alpha$0.600$1.808,192BudgetEU-sovereign multilingual instruct model — GDPR + BSI C5 compliant
GLM-4.6Zhipu$0.600$2.20200,000FlagshipZhipu flagship reasoning and coding model
Mistral Large 4Mistral$0.600$2.40256,000FlagshipNext-gen European flagship — multilingual and function calling
MiniMax M4MiniMax$0.600$2.404,000,000PremiumSuccessor to M3 — 4M context with stronger agentic reasoning
Kimi CoderMoonshot$0.650$2.60262,144FlagshipRepo-scale code generation, refactoring, and agentic dev
GLM-4-PlusZhipu$0.700$2.10128,000FlagshipPremium GLM variant for complex reasoning and tool use
MiMo V3Xiaomi$0.700$2.10512,000FlagshipNext-gen open-weight MoE — long-context reasoning and tool use
Kimi MathMoonshot$0.700$2.80128,000Mid-tierSpecialized math and theorem-proving variant
DeepSeek R2DeepSeek$0.700$2.80256,000ReasoningSuccessor to R1 — frontier open-weight reasoning at budget prices
Qwen 2.5 Coder 32B (Together AI)Together AI$0.800$0.800131,072Mid-tierHosted Qwen 2.5 Coder 32B for IDE-grade completions
Reka FlashReka$0.800$2.00128,000Mid-tierFast multimodal model for production agents
MiMo ProXiaomi$0.800$2.40256,000FlagshipProduction MoE flagship with extended context and tool use
Llama 4 BehemothMeta$0.800$2.40256,000Premium2T-parameter teacher model — frontier open-weight quality
Nemotron 4 UltraNvidia$0.800$2.40256,000FlagshipFrontier Nemotron MoE for agentic reasoning and tool use
GLM-5Zhipu$0.800$3.00256,000FlagshipNext-gen GLM flagship for reasoning, coding, and tool use
Nova ProAmazon$0.800$3.20300,000FlagshipHigh-quality text, vision, and tool use via Amazon Bedrock
Kimi K1.5 LongMoonshot$0.800$3.202,000,000Flagship2M-token context for full codebases and document corpora
Kimi AudioMoonshot$0.800$3.2064,000Mid-tierSpeech-in / speech-out multimodal audio model
o5-miniOpenAI$0.800$3.20256,000ReasoningCompact o5 reasoning for everyday problem solving
Claude 3.5 HaikuAnthropic$0.800$4.00200,000BudgetFast, affordable Claude 3.5 for chat and classification
Llama 3.3 70B (Cerebras)Cerebras$0.850$1.20128,000Mid-tierWafer-scale Llama 3.3 70B — 2000+ tokens/sec inference
Llama 3.3 70B (Together AI)Together AI$0.880$0.880128,000Mid-tierHosted Llama 3.3 70B with fast turbo endpoints
DeepSeek V3 (Together AI)Together AI$0.900$0.900128,000BudgetHosted DeepSeek V3 with fast autoscaling inference
Llama 3.3 70B (Fireworks AI)Fireworks AI$0.900$0.900128,000Mid-tierFireworks-hosted Llama 3.3 70B — low-latency inference
Qwen 2.5 Coder 32B (Fireworks AI)Fireworks AI$0.900$0.900131,072Mid-tierFireworks-hosted Qwen 2.5 Coder 32B for agentic dev tools
MiniMax ProMiniMax$0.900$2.70256,000FlagshipMax-capability variant with extended-context reasoning and tool use
Hy3 ProTencent$0.900$2.70256,000FlagshipProduction-grade MoE flagship for complex agents and long-context RAG
Kimi VisionMoonshot$0.900$3.60128,000FlagshipMultimodal vision-language model for charts, OCR, and UI
Kimi K2.6Moonshot$0.950$4.00262,144FlagshipFrontier-scale coding, long-context agentic workflows
Mistral OCRMistral$1.00$1.000Mid-tierDocument OCR with layout and table extraction
SonarPerplexity$1.00$1.00127,000BudgetFast grounded web search and everyday Q&A with citations
Grok VoicexAI$1.00$4.0032,000Mid-tierRealtime voice-mode model for spoken interactions
Claude Haiku 4.5Anthropic$1.00$5.00200,000BudgetClassification, chat, extraction
Qwen3 CoderQwen$1.00$5.001,000,000FlagshipBest-in-class open-weight coding and tool use
Sonar ReasoningPerplexity$1.00$5.00127,000ReasoningChain-of-thought reasoning grounded in real-time web search
Grok 1.5xAI$1.00$5.00128,000Mid-tierOriginal Grok 1.5 — extended-context chat and reasoning
Claude Haiku 4.6Anthropic$1.00$5.00200,000BudgetFaster Haiku refresh — chat, classification, extraction
Speechmatics EnhancedSpeechmatics$1.04$00FlagshipBest-in-class STT: $1.04 per audio hour, on-prem available
o4-miniOpenAI$1.10$4.40128,000ReasoningStrong problem-solving and logical deduction
o3-miniOpenAI$1.10$4.40128,000ReasoningStrong problem-solving and logical deduction
MiMo V2.5 ProXiaomi$1.20$3.60256,000FlagshipFlagship MiMo with extended context and agentic tool use
Mixtral 8x22B (Together AI)Together AI$1.20$3.6064,000FlagshipHosted sparse MoE — multilingual and long-context
Kimi K2 ThinkingMoonshot$1.20$5.00262,144PremiumDeep step-by-step reasoning across math, science, and code
Kimi K3Moonshot$1.20$5.00512,000PremiumNext-gen Kimi flagship — agentic coding, 512K context, tool use
Gemini 1.5 ProGoogle$1.25$5.002,000,000FlagshipPrevious-gen long-context multimodal flagship
Gemini 2.5 ProGoogle$1.25$10.002,000,000FlagshipCoding, analysis, long context
GPT-5OpenAI$1.25$10.00400,000FlagshipPowerful all-rounder for accuracy, nuance, and multimodal text+image tasks
GLM-5.2Z.AI$1.40$4.201,000,000FlagshipZ.ai 744B open flagship — #1 open on coding & agent benchmarks, MIT license, 1M context
GLM-5.2Zhipu$1.40$4.201,000,000FlagshipZ.ai 744B open flagship — #1 open on coding & agent benchmarks, MIT license, 1M context
Kimi ResearcherMoonshot$1.50$6.00262,144PremiumAutonomous deep-research agent with browsing and citations
Qwen3 MaxQwen$1.60$6.401,000,000FlagshipFrontier reasoning, agentic coding, long context
GPT-5.2OpenAI$1.75$14.00128,000FlagshipCoding, analysis
Mixtral 8x22B (Fireworks AI)Fireworks AI$1.80$5.4064,000FlagshipFireworks-hosted Mixtral 8x22B with speculative decoding
Rerank 3Cohere$2.00$04,096Mid-tierCross-encoder reranking for high-precision RAG ($2 per 1K searches)
Reka CoreReka$2.00$2.00128,000FlagshipMultimodal flagship across text, image, audio, and video
Magistral Medium 2Mistral$2.00$5.00128,000PremiumFrontier European reasoning model with extended chain-of-thought
Pixtral LargeMistral$2.00$6.00128,000FlagshipMultimodal flagship — vision, charts, document understanding
Mixtral 8x22BMistral$2.00$6.0064,000Open SourceSparse mixture-of-experts open-weight model
Hy3 MaxTencent$2.00$6.00256,000PremiumMaximum-capability Hy3 with extended-context reasoning and tool use
Mistral Large 2Mistral$2.00$6.00128,000FlagshipPrevious-gen Mistral flagship — strong reasoning and multilingual
GPT-4.1OpenAI$2.00$8.001,000,000FlagshipCoding, analysis, long-context tasks
o3OpenAI$2.00$8.00200,000ReasoningComplex reasoning, research, math
Sonar Reasoning ProPerplexity$2.00$8.00127,000ReasoningAdvanced DeepSeek R1-based reasoning with grounded search
Sonar Deep ResearchPerplexity$2.00$8.00127,000PremiumExpert multi-query research, long-form reports with citations
Jamba 1.7 LargeAI21$2.00$8.00256,000FlagshipHybrid Mamba-Transformer flagship with 256K context
GPT-Image 1 MiniOpenAI$2.00$8.000Mid-tierCost-efficient image generation via /v1/images/generations
Qwen3.5 MaxQwen$2.00$8.001,000,000PremiumNext-gen Qwen flagship — frontier reasoning and agentic coding
Grok 4.1xAI$2.00$10.002,000,000FlagshipLong-context flagship with native tool use and search
Grok 2 VisionxAI$2.00$10.0032,768Mid-tierMultimodal Grok 2 — charts, documents, and screenshots
Grok 2xAI$2.00$10.00128,000FlagshipPrevious-gen Grok with strong reasoning and search
Gemini 3.1 ProGoogle$2.00$12.001,000,000PremiumLong context, multimodal
Gemini 3 ProGoogle$2.00$12.002,000,000PremiumFrontier multimodal, agentic capabilities
Reka Vision ProReka$2.50$5.00256,000PremiumPremium multimodal — video, image, audio, and text reasoning
Command ACohere$2.50$10.00256,000FlagshipEnterprise agents, RAG, and tool use
Command R+Cohere$2.50$10.00128,000FlagshipAdvanced RAG and multilingual workflows
GPT-4o (GitHub Models)GitHub Models$2.50$10.00128,000FlagshipAzure-hosted GPT-4o via GitHub Models
GPT-4o (via Vercel)Vercel AI Gateway$2.50$10.00128,000FlagshipRouted pass-through pricing
Nova PremierAmazon$2.50$12.501,000,000PremiumLargest Bedrock-native model — frontier multimodal reasoning
GPT-5.4OpenAI$2.50$15.00128,000FlagshipGeneral purpose, agents, coding
Gemini 3 Pro VisionGoogle$2.50$15.002,000,000PremiumVision-specialized Gemini 3 Pro for charts, documents, and video
Llama 3.1 405BMeta$2.70$2.70128,000Open SourceLargest open-weight Llama — frontier-class quality
Yi-Large01.AI$3.00$3.0032,768FlagshipBilingual EN/ZH reasoning model
DeepSeek R1 (Together AI)Together AI$3.00$7.00128,000ReasoningHosted DeepSeek R1 reasoning with autoscaling inference
DeepSeek R1 (Fireworks AI)Fireworks AI$3.00$8.00128,000ReasoningFireworks-hosted DeepSeek R1 with optimized throughput
Hy3 UltraTencent$3.00$9.00512,000PremiumLargest Hy3 variant — frontier reasoning and extended context
o5OpenAI$3.00$12.00256,000ReasoningNext-gen o-series reasoning — math, science, agentic planning
Command A ReasoningCohere$3.00$12.00256,000ReasoningReasoning-tuned Command A for enterprise agents and analysis
Claude Sonnet 4.6Anthropic$3.00$15.00200,000FlagshipProduction workhorse, instruction following
Claude Sonnet 4.5Anthropic$3.00$15.00200,000FlagshipRecommended default, strong coding
Grok 4xAI$3.00$15.00256,000FlagshipFrontier reasoning and coding
Grok 3xAI$3.00$15.00131,072FlagshipPrevious-gen flagship — strong general reasoning
Sonar ProPerplexity$3.00$15.00200,000FlagshipMulti-step research with 2× more citations and deeper sources
Claude 3.5 Sonnet (via Vercel)Vercel AI Gateway$3.00$15.00200,000FlagshipRouted via Vercel AI Gateway
Claude 3 SonnetAnthropic$3.00$15.00200,000Mid-tierOriginal Claude 3 Sonnet — balanced reasoning and cost
Claude 3.5 SonnetAnthropic$3.00$15.00200,000FlagshipPrevious-gen workhorse — coding, vision, and tool use
Claude Sonnet 4.7Anthropic$3.00$15.001,000,000FlagshipLatest Sonnet — 1M context, stronger coding and tool use
Claude Sonnet 3.7Anthropic$3.00$15.00200,000FlagshipHybrid reasoning Sonnet 3.7 with extended thinking mode
Grok 5xAI$4.00$20.002,000,000PremiumNext-gen Grok flagship — agentic reasoning and native search
Hume EVI 3Hume AI$4.32$00FlagshipEmpathic voice (STT+LLM+TTS): $0.072/min ≈ $4.32 per audio hour
GPT-4o RealtimeOpenAI$5.00$20.00128,000FlagshipLow-latency speech-to-speech Realtime API model
Sonar UltraPerplexity$5.00$20.00200,000PremiumLargest Sonar variant — multi-step research with extensive citations
Claude Opus 4.8Anthropic$5.00$25.001,000,000PremiumPrevious Anthropic flagship — frontier reasoning, agentic coding, 1M context
Claude Opus 4.7Anthropic$5.00$25.001,000,000PremiumFrontier reasoning, agentic coding, long-context analysis
Claude Opus 4.6Anthropic$5.00$25.001,000,000PremiumDeep reasoning, long documents, coding
Claude Opus 4.5Anthropic$5.00$25.00200,000PremiumPrevious-gen flagship, stable versioned
Grok 3 ReasoningxAI$5.00$25.00131,072PremiumExtended chain-of-thought reasoning variant of Grok 3
Gemini 3.1 UltraGoogle$5.00$25.002,000,000PremiumLargest-scale Gemini 3.1 for frontier reasoning and agentic workflows
GPT-5.5OpenAI$5.00$30.00256,000FlagshipMost intuitive flagship — complex multi-step tasks, agents
Grok 4 HeavyxAI$6.00$30.00256,000PremiumMulti-agent parallel reasoning — top scores on ARC-AGI and HLE
GPT-5.6OpenAI$6.00$36.00400,000FlagshipNext-gen flagship — deeper reasoning, multimodal, long-context agents
GPT-Image 2OpenAI$10.00$40.000PremiumState-of-the-art image generation via /v1/images/generations
Claude Fable 5Anthropic$10.00$50.001,000,000PremiumMythos-class flagship — SOTA coding, vision, long-horizon agentic work
Octave TTSHume AI$15.00$05,000Mid-tierEmotionally expressive TTS (~$0.015/1K chars)
Claude 3 OpusAnthropic$15.00$75.00200,000PremiumOriginal Opus flagship — complex reasoning and writing
GPT-5.5 ProOpenAI$30.00$180.00256,000PremiumFrontier reasoning, research, agentic workflows
GPT-5.4 ProOpenAI$30.00$180.00256,000PremiumHighly complex tasks and cutting-edge research
Eleven Flash v2.5ElevenLabs$33.00$05,000BudgetCheapest TTS (~$0.033/1K chars)
Luminous SupremeAleph Alpha$35.00$35.002,048PremiumPremium multilingual completions on EU-sovereign infrastructure
Eleven Turbo v2.5ElevenLabs$50.00$05,000Mid-tierLow-latency TTS (~$0.05/1K chars)
MiniMax Speech-2.5 HDMiniMax$60.00$05,000FlagshipMultilingual expressive TTS (~$0.06/1K chars) — 30+ languages
GPT-4.5OpenAI$75.00$150.00128,000PremiumOriginal GPT-4.5 research preview — broad knowledge and writing
Eleven Multilingual v2ElevenLabs$180.00$05,000PremiumTTS billed per character (~$0.18/1K chars)