LLM cost calculator
Describe the requests you actually send and every model is priced against them, cheapest first. Output typically costs several times more than input, so the cheapest model for a summariser is rarely the cheapest for a chat agent — which a per-token price list cannot tell you.
Start from a typical workload
Prompt, system message and any retrieved context
What the model generates back
Total calls across all users
Share of the prompt that repeats between calls. Usually the single largest saving available.
$6.30
Cheapest — GLM-OCR
$58.5K
Most expensive — o1-pro
9286×
Difference between them
3 models would cost nothing
Listed separately rather than at the top of the ranking, where they would win every comparison by default and tell you nothing: GLM-4.5-Flash, GLM-4.6V-Flash, GLM-4.7-Flash.
Monthly cost, cheapest first
195 paid models priced for 100,000 requests of 1,500 in / 600 out.
| # | Model | Provider | Per request | Per month | Context |
|---|---|---|---|---|---|
| 1 | GLM-OCR | Zhipu AI (GLM) | $0.0001 | $6.30 | — |
| 2 | Qwen3.7 FlashVia OpenRouter | Alibaba (Qwen) | $0.0001 | $11.22 | 1M |
| 3 | DeepSeek V4 Flash LatestVia OpenRouter | DeepSeek | $0.0002 | $16.85 | 1M |
| 4 | Qwen3 30B A3B Instruct 2507Via OpenRouter | Alibaba (Qwen) | $0.0002 | $18.81 | 262K |
| 5 | DeepSeek V4 Flash 0731Via OpenRouter | DeepSeek | $0.0002 | $19.92 | 1M |
| 6 | GLM-4-32B-0414-128K | Zhipu AI (GLM) | $0.0002 | $21.00 | — |
| 7 | Qwen3.5-9BVia OpenRouter | Alibaba (Qwen) | $0.0002 | $24.00 | 262K |
| 8 | Qwen3.5-FlashVia OpenRouter | Alibaba (Qwen) | $0.0003 | $25.35 | 1M |
| 9 | Qwen2.5 7B InstructVia OpenRouter | Alibaba (Qwen) | $0.0003 | $27.00 | 33K |
| 10 | UI-TARS 7B Via OpenRouter | ByteDance (Doubao) | $0.0003 | $27.00 | 128K |
| 11 | Qwen3 Coder 30B A3B InstructVia OpenRouter | Alibaba (Qwen) | $0.0003 | $27.30 | 262K |
| 12 | GLM-4.6V-FlashX | Zhipu AI (GLM) | $0.0003 | $28.38 | — |
| 13 | Qwen3 32BVia OpenRouter | Alibaba (Qwen) | $0.0003 | $28.80 | 131K |
| 14 | Gemini 2.0 Flash-Lite | $0.0003 | $29.25 | — | |
| 15 | gpt-5-nano | OpenAI | $0.0003 | $29.47 | 400K |
| 16 | GLM-4.7-FlashX | Zhipu AI (GLM) | $0.0003 | $31.80 | — |
| 17 | Qwen3 14BVia OpenRouter | Alibaba (Qwen) | $0.0003 | $32.40 | 131K |
| 18 | DeepSeek V4 Flash 0423Via OpenRouter | DeepSeek | $0.0003 | $32.76 | 1M |
| 19 | Gemini 2.5 Flash-Lite | $0.0003 | $34.95 | 1M | |
| 20 | Gemini 2.5 Flash-Lite Preview | $0.0003 | $34.95 | — | |
| 21 | Gemini 2.0 Flash | $0.0004 | $35.63 | — | |
| 22 | gpt-4.1-nano | OpenAI | $0.0004 | $35.63 | 1M |
| 23 | Qwen3 VL 32B InstructVia OpenRouter | Alibaba (Qwen) | $0.0004 | $40.56 | 131K |
| 24 | Qwen3 8BVia OpenRouter | Alibaba (Qwen) | $0.0004 | $44.85 | 131K |
| 25 | Qwen3 VL 8B InstructVia OpenRouter | Alibaba (Qwen) | $0.0004 | $44.85 | 262K |
| 26 | Qwen3 235B A22B Instruct 2507Via OpenRouter | Alibaba (Qwen) | $0.0005 | $46.50 | 262K |
| 27 | Gemini 2.5 Flash Image (Nano Banana) 🍌 | $0.0005 | $47.34 | 33K | |
| 28 | Qwen3 30B A3BVia OpenRouter | Alibaba (Qwen) | $0.0005 | $48.00 | 131K |
| 29 | gpt-4o-mini | OpenAI | $0.0006 | $55.13 | 128K |
| 30 | DeepSeek V3.2Via OpenRouter | DeepSeek | $0.0006 | $58.30 | 164K |
| 31 | Qwen3 VL 30B A3B InstructVia OpenRouter | Alibaba (Qwen) | $0.0006 | $58.50 | 262K |
| 32 | Qwen3 Coder NextVia OpenRouter | Alibaba (Qwen) | $0.0006 | $63.75 | 262K |
| 33 | DeepSeek V3.2 ExpVia OpenRouter | DeepSeek | $0.0007 | $65.10 | 164K |
| 34 | gpt-4.1-nano-2025-04-14 | OpenAI | $0.0007 | $71.25 | — |
| 35 | Qwen-PlusVia OpenRouter | Alibaba (Qwen) | $0.0008 | $76.44 | 1M |
| 36 | Qwen3.6 35B A3BVia OpenRouter | Alibaba (Qwen) | $0.0008 | $78.00 | 262K |
| 37 | Qwen2.5 72B InstructVia OpenRouter | Alibaba (Qwen) | $0.0008 | $78.00 | 33K |
| 38 | Qwen3 Next 80B A3B InstructVia OpenRouter | Alibaba (Qwen) | $0.0008 | $79.50 | 262K |
| 39 | Qwen3 Coder FlashVia OpenRouter | Alibaba (Qwen) | $0.0008 | $80.73 | 1M |
| 40 | Qwen3.5-35B-A3BVia OpenRouter | Alibaba (Qwen) | $0.0008 | $81.00 | 262K |
| 41 | Qwen2.5 VL 72B InstructVia OpenRouter | Alibaba (Qwen) | $0.0008 | $82.50 | 128K |
| 42 | babbage-002 | OpenAI | $0.0008 | $84.00 | — |
| 43 | Qwen Plus 0728Via OpenRouter | Alibaba (Qwen) | $0.0009 | $85.80 | 1M |
| 44 | GLM-4.6V | Zhipu AI (GLM) | $0.0009 | $87.75 | 131K |
| 45 | GLM-4.5-Air | Zhipu AI (GLM) | $0.0009 | $88.35 | 131K |
| 46 | DeepSeek V3.1Via OpenRouter | DeepSeek | $0.0009 | $89.10 | 164K |
| 47 | gpt-5.6-luna | OpenAI | $0.0009 | $93.90 | 1.1M |
| 48 | DeepSeek V3.1 TerminusVia OpenRouter | DeepSeek | $0.0009 | $94.42 | 164K |
| 49 | Qwen3 Next 80B A3B ThinkingVia OpenRouter | Alibaba (Qwen) | $0.0009 | $94.50 | 262K |
| 50 | Qwen3.6 FlashVia OpenRouter | Alibaba (Qwen) | $0.0010 | $95.63 | 1M |
| 51 | Qwen3 Coder 480B A35BVia OpenRouter | Alibaba (Qwen) | $0.0010 | $96.00 | 262K |
| 52 | gpt-5.4-nano | OpenAI | $0.0010 | $96.90 | 400K |
| 53 | DeepSeek V3Via OpenRouter | DeepSeek | $0.0010 | $100.33 | 164K |
| 54 | DeepSeek V3 0324Via OpenRouter | DeepSeek | $0.0010 | $101.63 | 164K |
| 55 | gpt-4o-mini-2024-07-18 | OpenAI | $0.0011 | $110.25 | 128K |
| 56 | Qwen3.7 PlusVia OpenRouter | Alibaba (Qwen) | $0.0011 | $113.28 | 1M |
| 57 | Gemini 3.1 Flash-Lite | $0.0012 | $117.38 | 1M | |
| 58 | Qwen3.5-27BVia OpenRouter | Alibaba (Qwen) | $0.0012 | $122.85 | 262K |
| 59 | Gemini 3.1 Flash Lite Image (Nano Banana 2 Lite) 🍌 | $0.0013 | $127.50 | 66K | |
| 60 | Qwen Plus 0728 (thinking)Via OpenRouter | Alibaba (Qwen) | $0.0013 | $132.00 | 1M |
| 61 | Qwen3.5 Plus 2026-02-15Via OpenRouter | Alibaba (Qwen) | $0.0013 | $132.60 | 1M |
| 62 | ERNIE 4.5 VL 424B A47B Via OpenRouter | Baidu (ERNIE) | $0.0014 | $138.00 | 123K |
| 63 | Qwen3 VL 235B A22B InstructVia OpenRouter | Alibaba (Qwen) | $0.0014 | $140.55 | 262K |
| 64 | gpt-4.1-mini | OpenAI | $0.0014 | $142.50 | 1M |
| 65 | DeepSeek V4 ProVia OpenRouter | DeepSeek | $0.0014 | $144.53 | 1M |
| 66 | gpt-5-mini | OpenAI | $0.0015 | $147.38 | 400K |
| 67 | Qwen3 VL 8B ThinkingVia OpenRouter | Alibaba (Qwen) | $0.0015 | $153.00 | 131K |
| 68 | Qwen3.5 Plus 2026-04-20Via OpenRouter | Alibaba (Qwen) | $0.0015 | $153.00 | 1M |
| 69 | Qwen2.5 Coder 32B InstructVia OpenRouter | Alibaba (Qwen) | $0.0016 | $159.00 | 33K |
| 70 | gpt-3.5-turbo | OpenAI | $0.0016 | $165.00 | 16K |
| 71 | gpt-3.5-turbo-0125 | OpenAI | $0.0016 | $165.00 | — |
| 72 | Qwen3.6 PlusVia OpenRouter | Alibaba (Qwen) | $0.0017 | $165.75 | 1M |
| 73 | R1 Distill Llama 70BVia OpenRouter | DeepSeek | $0.0017 | $168.00 | 8K |
| 74 | Qwen3 235B A22B Thinking 2507Via OpenRouter | Alibaba (Qwen) | $0.0017 | $172.50 | 262K |
| 75 | Qwen3 30B A3B Thinking 2507Via OpenRouter | Alibaba (Qwen) | $0.0017 | $174.00 | 82K |
| 76 | Qwen3 VL 30B A3B ThinkingVia OpenRouter | Alibaba (Qwen) | $0.0017 | $174.00 | 262K |
| 77 | Kimi K2.7 Code (batch)Via OpenRouter | Moonshot AI (Kimi) | $0.0017 | $174.15 | 262K |
| 78 | GLM-4.5V | Zhipu AI (GLM) | $0.0018 | $175.95 | 66K |
| 79 | Qwen3 235B A22BVia OpenRouter | Alibaba (Qwen) | $0.0018 | $177.45 | 131K |
| 80 | Gemini 2.5 Flash | $0.0018 | $182.85 | 1M | |
| 81 | Gemini 3.5 Flash-Lite | $0.0018 | $182.85 | 1M | |
| 82 | Qwen3.5-122B-A10BVia OpenRouter | Alibaba (Qwen) | $0.0019 | $187.50 | 262K |
| 83 | Gemini 2.5 Flash Native Audio (Live API) | $0.0019 | $195.00 | — | |
| 84 | R1 0528Via OpenRouter | DeepSeek | $0.0020 | $197.25 | 164K |
| 85 | GLM-4.5 | Zhipu AI (GLM) | $0.0020 | $199.95 | 131K |
| 86 | GLM-4.6 | Zhipu AI (GLM) | $0.0020 | $199.95 | 205K |
| 87 | GLM-4.7 | Zhipu AI (GLM) | $0.0020 | $199.95 | 205K |
| 88 | Kimi K2.6Via OpenRouter | Moonshot AI (Kimi) | $0.0021 | $211.64 | 262K |
| 89 | Kimi K2 ThinkingVia OpenRouter | Moonshot AI (Kimi) | $0.0022 | $219.75 | 262K |
| 90 | Kimi K2 0711Via OpenRouter | Moonshot AI (Kimi) | $0.0022 | $223.50 | 131K |
| 91 | grok-build-0.1 | xAI | $0.0023 | $234.00 | 256K |
| 92 | Gemini 3 Flash Preview | $0.0023 | $234.75 | 1M | |
| 93 | Kimi K2.5Via OpenRouter | Moonshot AI (Kimi) | $0.0024 | $235.13 | 262K |
| 94 | Kimi K2 0905Via OpenRouter | Moonshot AI (Kimi) | $0.0024 | $240.00 | 262K |
| 95 | Gemini 3.1 Flash Image (Nano Banana 2) 🍌 | $0.0026 | $255.00 | 131K | |
| 96 | R1Via OpenRouter | DeepSeek | $0.0026 | $255.00 | 164K |
| 97 | Qwen3 Coder PlusVia OpenRouter | Alibaba (Qwen) | $0.0027 | $269.10 | 1M |
| 98 | gpt-3.5-turbo-1106 | OpenAI | $0.0027 | $270.00 | — |
| 99 | Qwen3.5 397B A17BVia OpenRouter | Alibaba (Qwen) | $0.0028 | $282.00 | 262K |
| 100 | Qwen3.6 27BVia OpenRouter | Alibaba (Qwen) | $0.0028 | $284.40 | 262K |
| 101 | gpt-4.1-mini-2025-04-14 | OpenAI | $0.0029 | $285.00 | — |
| 102 | Kimi K2.7 CodeVia OpenRouter | Moonshot AI (Kimi) | $0.0029 | $290.25 | 262K |
| 103 | grok-4.20-0309-non-reasoning | xAI | $0.0029 | $290.25 | 1M |
| 104 | grok-4.20-0309-reasoning | xAI | $0.0029 | $290.25 | 1M |
| 105 | grok-4.20-multi-agent-0309 | xAI | $0.0029 | $290.25 | 1M |
| 106 | grok-4.3 | xAI | $0.0029 | $290.25 | 1M |
| 107 | Qwen3 VL 235B A22B ThinkingVia OpenRouter | Alibaba (Qwen) | $0.0030 | $300.00 | 131K |
| 108 | GLM-5 | Zhipu AI (GLM) | $0.0031 | $306.00 | 205K |
| 109 | Qwen3 MaxVia OpenRouter | Alibaba (Qwen) | $0.0032 | $322.92 | 262K |
| 110 | Claude Haiku 3.5 | Anthropic | $0.0033 | $327.60 | — |
| 111 | gpt-3.5-turbo-instruct | OpenAI | $0.0034 | $345.00 | 4K |
| 112 | Qwen3 Max ThinkingVia OpenRouter | Alibaba (Qwen) | $0.0035 | $351.00 | 262K |
| 113 | gpt-5.4-mini | OpenAI | $0.0035 | $352.13 | 400K |
| 114 | GLM-5-Turbo | Zhipu AI (GLM) | $0.0038 | $376.80 | 203K |
| 115 | GLM-5V-Turbo | Zhipu AI (GLM) | $0.0038 | $376.80 | 203K |
| 116 | Gemini 3.1 Flash Live Preview | $0.0038 | $382.50 | — | |
| 117 | o4-mini | OpenAI | $0.0039 | $391.88 | 200K |
| 118 | GLM-4.5-AirX | Zhipu AI (GLM) | $0.0040 | $395.40 | — |
| 119 | o3-mini | OpenAI | $0.0040 | $404.25 | 200K |
| 120 | Claude Haiku 4.5 | Anthropic | $0.0041 | $409.50 | 200K |
| 121 | davinci-002 | OpenAI | $0.0042 | $420.00 | — |
| 122 | GLM-5.1 | Zhipu AI (GLM) | $0.0042 | $422.70 | 205K |
| 123 | GLM-5.2 | Zhipu AI (GLM) | $0.0042 | $422.70 | 1M |
| 124 | Qwen3.7 MaxVia OpenRouter | Alibaba (Qwen) | $0.0043 | $433.65 | 1M |
| 125 | Gemini Robotics ER 1.6 Preview | $0.0045 | $450.00 | — | |
| 126 | gpt-4o-mini-transcribe | OpenAI | $0.0049 | $487.50 | — |
| 127 | Qwen3.6 Max PreviewVia OpenRouter | Alibaba (Qwen) | $0.0052 | $523.77 | 262K |
| 128 | Qwen3.8 MaxVia OpenRouter | Alibaba (Qwen) | $0.0058 | $581.25 | 1M |
| 129 | grok-4.5 | xAI | $0.0058 | $583.50 | 500K |
| 130 | Gemini 3.6 Flash | $0.0061 | $614.25 | 1M | |
| 131 | Gemini 2.5 Flash Preview TTS | $0.0067 | $675.00 | — | |
| 132 | Gemini 3.5 Flash | $0.0070 | $704.25 | 1M | |
| 133 | gpt-4.1 | OpenAI | $0.0071 | $712.50 | 1M |
| 134 | o3 | OpenAI | $0.0071 | $712.50 | 200K |
| 135 | Gemini 2.5 Pro | $0.0074 | $736.88 | 1M | |
| 136 | gpt-5 | OpenAI | $0.0074 | $736.88 | 400K |
| 137 | gpt-5-search-api | OpenAI | $0.0074 | $736.88 | — |
| 138 | gpt-5.1 | OpenAI | $0.0074 | $736.88 | 400K |
| 139 | gpt-image-1-mini | OpenAI | $0.0075 | $753.75 | — |
| 140 | Gemini Omni Flash Preview | $0.0076 | $765.00 | — | |
| 141 | GLM-4.5-X | Zhipu AI (GLM) | $0.0079 | $785.25 | — |
| 142 | Gemini 2.5 Computer Use Preview | $0.0079 | $787.50 | — | |
| 143 | Claude Sonnet 5 | Anthropic | $0.0082 | $819.00 | 1M |
| 144 | Gemini Robotics ER 2 Preview | $0.0082 | $819.00 | — | |
| 145 | Gemini Robotics ER 2 Streaming Preview | $0.0090 | $900.00 | — | |
| 146 | gpt-4o | OpenAI | $0.0092 | $918.75 | 128K |
| 147 | Gemini 3.1 Pro Preview | $0.0094 | $939.00 | 1M | |
| 148 | gpt-5.6-terra | OpenAI | $0.0094 | $939.00 | 1.1M |
| 149 | gpt-4o-transcribe | OpenAI | $0.0097 | $975.00 | — |
| 150 | gpt-4o-transcribe-diarize | OpenAI | $0.0097 | $975.00 | — |
| 151 | Gemini 3 Pro Image (Nano Banana Pro) 🍌 | $0.01 | $1.0K | 131K | |
| 152 | gpt-5.2 | OpenAI | $0.01 | $1.0K | 400K |
| 153 | gpt-5.2-chat-latest | OpenAI | $0.01 | $1.0K | — |
| 154 | gpt-5.3-chat-latest | OpenAI | $0.01 | $1.0K | — |
| 155 | gpt-5.3-codex | OpenAI | $0.01 | $1.0K | 400K |
| 156 | gpt-4.1-2025-04-14 | OpenAI | $0.01 | $1.1K | — |
| 157 | MoonshotAI Kimi LatestVia OpenRouter | Moonshot AI (Kimi) | $0.01 | $1.1K | 1M |
| 158 | gpt-5.4 | OpenAI | $0.01 | $1.2K | 1.1M |
| 159 | Claude Sonnet 4 | Anthropic | $0.01 | $1.2K | 1M |
| 160 | Claude Sonnet 4.5 | Anthropic | $0.01 | $1.2K | 1M |
| 161 | Claude Sonnet 4.6 | Anthropic | $0.01 | $1.2K | 1M |
| 162 | Kimi K3Via OpenRouter | Moonshot AI (Kimi) | $0.01 | $1.2K | 1M |
| 163 | Gemini 2.5 Pro Preview TTS | $0.01 | $1.4K | — | |
| 164 | Gemini 3.1 Flash TTS Preview | $0.01 | $1.4K | — | |
| 165 | gpt-4o-2024-08-06 | OpenAI | $0.01 | $1.4K | 128K |
| 166 | o4-mini-2025-04-16 | OpenAI | $0.01 | $1.4K | — |
| 167 | gpt-4o-2024-05-13 | OpenAI | $0.02 | $1.6K | 128K |
| 168 | Gemini 3.5 Live Translate | $0.02 | $1.8K | — | |
| 169 | Claude Opus 4.5 | Anthropic | $0.02 | $2.0K | 200K |
| 170 | Claude Opus 4.6 | Anthropic | $0.02 | $2.0K | 1M |
| 171 | Claude Opus 4.7 | Anthropic | $0.02 | $2.0K | 1M |
| 172 | Claude Opus 4.8 | Anthropic | $0.02 | $2.0K | 1M |
| 173 | Claude Opus 5 | Anthropic | $0.02 | $2.0K | 1M |
| 174 | chat-latest | OpenAI | $0.02 | $2.3K | — |
| 175 | gpt-5.5 | OpenAI | $0.02 | $2.3K | 1.1M |
| 176 | gpt-5.6-sol | OpenAI | $0.02 | $2.3K | 1.1M |
| 177 | gpt-image-2 | OpenAI | $0.03 | $2.7K | — |
| 178 | chatgpt-image-latest | OpenAI | $0.03 | $2.9K | — |
| 179 | gpt-image-1.5 | OpenAI | $0.03 | $2.9K | — |
| 180 | gpt-4-turbo-2024-04-09 | OpenAI | $0.03 | $3.3K | — |
| 181 | gpt-image-1 | OpenAI | $0.04 | $3.6K | — |
| 182 | Claude Fable 5 | Anthropic | $0.04 | $4.1K | 1M |
| 183 | Claude Mythos 5 | Anthropic | $0.04 | $4.1K | — |
| 184 | o1 | OpenAI | $0.06 | $5.5K | 200K |
| 185 | gpt-5.5-cyber | OpenAI | $0.06 | $5.9K | — |
| 186 | gpt-5.6-cyber | OpenAI | $0.06 | $5.9K | — |
| 187 | Claude Opus 4 | Anthropic | $0.06 | $6.1K | 200K |
| 188 | Claude Opus 4.1 | Anthropic | $0.06 | $6.1K | 200K |
| 189 | o3-pro | OpenAI | $0.08 | $7.8K | 200K |
| 190 | gpt-4-0613 | OpenAI | $0.08 | $8.1K | — |
| 191 | gpt-5-pro | OpenAI | $0.09 | $9.4K | 400K |
| 192 | gpt-5.2-pro | OpenAI | $0.13 | $13.2K | 400K |
| 193 | gpt-5.4-pro | OpenAI | $0.15 | $15.3K | 1.1M |
| 194 | gpt-5.5-pro | OpenAI | $0.15 | $15.3K | 1.1M |
| 195 | o1-pro | OpenAI | $0.58 | $58.5K | 200K |
18 models excluded because they cannot serve this workload — mostly embedding, moderation and OCR endpoints that publish no output price because they do not generate text. Counting that missing price as zero would rank them as the cheapest way to run a chat.
Estimates use standard-tier list prices. They exclude batch discounts, committed-use agreements and free allowances, and assume every request is the same shape. Treat the ranking as sound and the absolute figures as approximate.
Questions about these estimates
- Why does the cheapest model change when I change the workload?
- Because output tokens usually cost three to five times what input tokens cost. A model with cheap input and expensive output wins for retrieval-style work with short answers and loses badly for anything that generates long text. Ranking on list price alone hides that.
- What does the cached input percentage do?
- Most providers bill a reduced rate — often around a tenth of the normal input price — when a prompt repeats a prefix they have already processed. If you send the same system prompt or document on every call, that share is billed at the cheaper rate. It is usually the single largest saving available.
- Why are some rows greyed out?
- Their context window is smaller than the tokens you entered, so they cannot hold your prompt. They are shown rather than hidden, because a model that cannot do the job is not a cheaper option.
- How accurate are these numbers?
- They use standard-tier list prices updated daily, and assume every request is the same shape. They exclude batch discounts, committed-use agreements and free allowances. Treat the ranking as sound and the absolute totals as an estimate.