10,634 prices on 4,757 listings · last collected 23 min ago · 399 flagged changes in 30 days
Every price, as each platform publishes it
A collector reads every public price on eight platforms, daily by default, and keeps the page it read. Nothing is typed in by hand. Search a model to see who charges least, per unit, right now.
Model
Claude 
Gemini 
Qwen 
DeepSeek 
GLM 
Kimi 
Mistral 
MiniMax 
Hunyuan 
Nemotron 
Llama Unit price
All 3,235Per second of output 13Per image 6Per million input tokens 1,107Per million output tokens 1,071Per million tokens 5Per million cached input tokens (read) 594Per million cache-write tokens 224Per thousand characters 2Per second of compute time 12Per run (default parameters) 193Other units (see the platform’s wording) 8
Showing 101–200 of 594
Per million cached input tokens (read)100
| Platform | Model · what it covers | USD | Platform’s own wording | Since |
|---|---|---|---|---|
| AibermNot disclosed | DeepSeek V3.2 | $0.021/M cache read | model_ratio 0.45 × $2/M · group default ×0.23 · cache_ratio 0.1 | |
| AibermNot disclosed | DeepSeek V3.2 Exp | $0.021/M cache read | model_ratio 0.45 × $2/M · group default ×0.23 · cache_ratio 0.1 | |
| AibermNot disclosed | DeepSeek V3.2 Exp Thinking | $0.021/M cache read | model_ratio 0.45 × $2/M · group default ×0.23 · cache_ratio 0.1 | |
| AibermNot disclosed | GPT-5 Mini | $0.021/M cache read | model_ratio 0.15 × $2/M · group default ×0.23 · cache_ratio 0.3 | |
| OpenRouterThird-party | DeepSeek: DeepSeek V4 Pro 0813via Baidu · fp8 | $0.021/M cache read | input_cache_read: 0.000000020958 / token (endpoint baidu/fp8) | |
| OpenRouterThird-party | Qwen: Qwen3.6 35B A3Bvia PrimeIntellect · fp8 | $0.021/M cache read | input_cache_read: 0.000000021 / token (endpoint primeintellect/fp8) | |
| OpenRouterOfficial | Tencent: Hy4 | $0.021/M cache read | input_cache_read: 0.000000021 / token (endpoint tencent/fp4) | |
| OpenRouterThird-party | DeepSeek: DeepSeek V3.2via GMICloud · fp8 | $0.022/M cache read | input_cache_read: 0.0000000216 / token (endpoint gmicloud/fp8) | |
| OpenRouterOfficial | Nex AGI: Nex-N2.6 | $0.024/M cache read | input_cache_read: 0.000000024 / token (endpoint nex-agi/fp8) | |
| OpenRouterOfficial | Anthropic: Claude Haiku 5.5 (batch)Prompt > 100K | $0.025/M cache read | input_cache_read: 0.000000025 / token (>100000; endpoint anthropic) | |
| OpenRouterOfficial | Google: Gemini 3.1 Flash Litevia Google AI Studio | $0.025/M cache read | input_cache_read: 0.000000025 / token (endpoint google-ai-studio) | |
| OpenRouterOfficial | Google: Gemini 3.1 Flash Lite Previewvia Google AI Studio | $0.025/M cache read | input_cache_read: 0.000000025 / token (endpoint google-ai-studio) | |
| OpenRouterOfficial | Google: Gemini 3.1 Flash Lite (batch) | $0.025/M cache read | input_audio_cache: 0.000000025 / token (endpoint google-vertex/global) | |
| OpenRouterOfficial | LiquidAI: LFM2.5-24B-A2B | $0.025/M cache read | input_cache_read: 0.000000025 / token (endpoint liquid/fp8) | |
| OpenRouterOfficial | OpenAI: GPT-6.1 Terra Pro (batch) | $0.025/M cache read | input_cache_read: 0.000000025 / token (endpoint openai) | |
| OpenRouterOfficial | OpenAI: GPT-6.1 Terra (batch) | $0.025/M cache read | input_cache_read: 0.000000025 / token (endpoint openai) | |
| OpenRouterThird-party | Qwen: Qwen3.6 27Bvia Chutes · fp8 | $0.025/M cache read | input_cache_read: 0.000000025 / token (endpoint chutes/fp8) | |
| OpenRouterThird-party | Ricursive: Ripple 65B Previewvia PrimeIntellect · bf16 | $0.025/M cache read | input_cache_read: 0.000000025 / token (endpoint primeintellect/bf16) | |
| OpenRouterThird-party | Z.ai: GLM 4.5 Airvia Novita · bf16 | $0.025/M cache read | input_cache_read: 0.000000025 / token (endpoint novita/bf16) | |
| AibermNot disclosed | GPT-5.1 Codex Max | $0.028/M cache read | model_ratio 0.6 × $2/M · group default ×0.23 · cache_ratio 0.1 | |
| AibermNot disclosed | DeepSeek V4 Flash | $0.029/M cache read | model_ratio 0.3126 × $2/M · group default ×0.23 · cache_ratio 0.2 | |
| KIENot disclosed | gpt-6-1-sol | $0.03/M cache read | 6 credits per million tokens ($0.03)Claims official or Fal price $0.1/M cache read | |
| OpenRouterOfficial | Google: Gemini 2.5 Flashvia Google AI Studio | $0.03/M cache read | input_cache_read: 0.00000003 / token (endpoint google-ai-studio) | |
| OpenRouterOfficial | Google: Gemini 2.5 Flash Litevia Google AI Studio | $0.03/M cache read | input_audio_cache: 0.00000003 / token (endpoint google-ai-studio) | |
| OpenRouterOfficial | Google: Gemini 3.5 Flash Litevia Google AI Studio | $0.03/M cache read | input_audio_cache: 0.00000003 / token (endpoint google-ai-studio) | |
| OpenRouterOfficial | Google: Gemini 3.5 Flash Litevia Google AI Studio | $0.03/M cache read | input_cache_read: 0.00000003 / token (endpoint google-ai-studio) | |
| OpenRouterOfficial | Inference.net: Schematron V2 Turbovia InferenceNet · bf16 | $0.03/M cache read | input_cache_read: 0.00000003 / token (endpoint inference-net/bf16) | |
| OpenRouterThird-party | MiniMax: MiniMax M2.5via Venice | $0.03/M cache read | input_cache_read: 0.00000003 / token (endpoint venice) | |
| OpenRouterOfficial | Mistral: Codestral 2508via Mistral | $0.03/M cache read | input_cache_read: 0.00000003 / token (endpoint mistral) | |
| OpenRouterOfficial partner | OpenAI: GPT-4.1 Nanovia Azure | $0.03/M cache read | input_cache_read: 0.00000003 / token (endpoint azure) | |
| OpenRouterOfficial partner | OpenAI: GPT-5 Minivia Azure | $0.03/M cache read | input_cache_read: 0.00000003 / token (endpoint azure) | |
| OpenRouterOfficial partner | OpenAI: GPT-5.1-Codex-Mini | $0.03/M cache read | input_cache_read: 0.00000003 / token | |
| OpenRouterOfficial | Pareto 26.10 | $0.03/M cache read | input_cache_read: 0.00000003 / token | |
| OpenRouterNot disclosed | Upstage: Solar Pro 4via Upstage · zdr | $0.03/M cache read | input_cache_read: 0.00000003 / token (endpoint upstage/zdr) | |
| OpenRouterNot disclosed | Z.ai: GLM 5.3 Flashvia Relace · fp4 | $0.03/M cache read | input_cache_read: 0.00000003 / token (endpoint relace/fp4) | |
| OpenRouterNot disclosed | Pareto Latest | $0.03/M cache read | input_cache_read: 0.00000003 / token | |
| OpenRouterNot disclosed | Z.ai: GLM Flash Latest | $0.03/M cache read | input_cache_read: 0.00000003 / token | |
| RunwareNot disclosed | Gemini 3.5 Flash-LitePer 1M | $0.03/M cache read | $3e-8 / cachedInputToken | |
| RunwareNot disclosed | GLM-5.3-FlashPer 1M | $0.03/M cache read | $3e-8 / cachedInputToken | |
| AibermNot disclosed | Claude Sonnet 5 | $0.03/M cache read | model_ratio 0.6522 × $2/M · group default ×0.23 · cache_ratio 0.1 | |
| OpenRouterThird-party | Meta: Muse Glimmer 30B (batch) | $0.032/M cache read | input_cache_read: 0.000000032 / token (endpoint deepinfra/bf16) | |
| OpenRouterThird-party | OpenAI: gpt-oss-120bvia PrimeIntellect · fp4 | $0.032/M cache read | input_cache_read: 0.000000032 / token (endpoint primeintellect/fp4) | |
| OpenRouterThird-party | Qwen: Qwen3.5-35B-A3Bvia Darkbloom · fp4 | $0.033/M cache read | input_cache_read: 0.0000000325 / token (endpoint darkbloom/fp4) | |
| OpenRouterThird-party | Tencent: Hy3peak 0000–1600 UTC · via Tencent · fp8 | $0.033/M cache read | input_cache_read: 0.000000033 / token (utc_start 0, utc_end 1600; endpoint tencent/fp8) | |
| OpenRouterNot disclosed | DeepSeek: DeepSeek V4 Flash Latest | $0.035/M cache read | input_cache_read: 0.000000035 / token | |
| RunwareNot disclosed | gpt-oss-safeguard-20bPer 1M | $0.035/M cache read | $3.5e-8 / cachedInputToken | |
| OpenRouterOfficial | Xiaomi: MiMo-V2.6-Pro-UltraSpeed | $0.036/M cache read | input_cache_read: 0.000000036 / token (endpoint xiaomi/fp8) | |
| OpenRouterOfficial | Google: Gemini 3.6 Flash (batch) | $0.038/M cache read | input_audio_cache: 0.0000000375 / token (endpoint google-vertex/global) | |
| OpenRouterOfficial | Google: Gemini 3.6 Flash (batch) | $0.038/M cache read | input_cache_read: 0.0000000375 / token (endpoint google-vertex/global) | |
| OpenRouterOfficial | Google: Gemini 3.7 Flash (batch) | $0.038/M cache read | input_audio_cache: 0.0000000375 / token (endpoint google-vertex/global) | |
| OpenRouterOfficial | Google: Gemini 3.7 Flash (batch) | $0.038/M cache read | input_cache_read: 0.0000000375 / token (endpoint google-vertex/global) | |
| OpenRouterOfficial | Google: Gemini 3.8 Flash (batch) | $0.038/M cache read | input_audio_cache: 0.0000000375 / token (endpoint google-vertex/global) | |
| OpenRouterOfficial | Google: Gemini 3.8 Flash (batch) | $0.038/M cache read | input_cache_read: 0.0000000375 / token (endpoint google-vertex/global) | |
| OpenRouterOfficial | OpenAI: GPT-4o-mini (batch) | $0.038/M cache read | input_cache_read: 0.0000000375 / token (endpoint openai) | |
| OpenRouterOfficial | OpenAI: GPT-5.4 Mini (batch) | $0.038/M cache read | input_cache_read: 0.0000000375 / token (endpoint openai) | |
| OpenRouterThird-party | OpenAI: gpt-oss-safeguard-20b | $0.038/M cache read | input_cache_read: 0.0000000375 / token | |
| OpenRouterThird-party | PrismML: Ternary Bonsai 2 27B | $0.038/M cache read | input_cache_read: 0.0000000375 / token (endpoint darkbloom/int4) | |
| OpenRouterThird-party | Tencent: Hy4 previewoff-peak 1600–0000 UTC · via Tencent · fp8 | $0.038/M cache read | input_cache_read: 0.0000000378 / token (utc_start 1600, utc_end 0; endpoint tencent/fp8) | |
| AibermNot disclosed | GPT-6 Sol | $0.039/M cache read | model_ratio 0.5652 × $2/M · group default ×0.23 · cache_ratio 0.15 | |
| AibermNot disclosed | GPT-6.1 Sol | $0.039/M cache read | model_ratio 0.5652 × $2/M · group default ×0.23 · cache_ratio 0.15 | |
| OpenRouterOfficial | Qwen: Qwen3 Coder Flash | $0.039/M cache read | input_cache_read: 0.000000039 / token (endpoint alibaba) | |
| AibermNot disclosed | Claude Opus 5.5 | $0.04/M cache read | model_ratio 0.8688 × $2/M · group default ×0.23 · cache_ratio 0.1 | |
| OpenRouterOfficial | KT: Mi:dm K 3.0 Pro | $0.04/M cache read | input_cache_read: 0.00000004 / token (endpoint kt) | |
| OpenRouterOfficial | LiquidAI: LFM2.5-2.6B | $0.04/M cache read | input_cache_read: 0.00000004 / token (endpoint liquid/fp8) | |
| OpenRouterThird-party | Meta: Muse Glimmer 30B | $0.04/M cache read | input_cache_read: 0.00000004 / token (endpoint deepinfra/bf16) | |
| OpenRouterOfficial | Mistral: Devstral 2 2512 | $0.04/M cache read | input_cache_read: 0.00000004 / token | |
| OpenRouterOfficial | Mistral: Mistral Medium 3via Mistral · zdr | $0.04/M cache read | input_cache_read: 0.00000004 / token (endpoint mistral/zdr) | |
| OpenRouterOfficial | Mistral: Mistral Medium 3.1via Mistral · zdr | $0.04/M cache read | input_cache_read: 0.00000004 / token (endpoint mistral/zdr) | |
| OpenRouterOfficial partner | OpenAI: GPT-5.6 LunaPrompt > 272Kvia Amazon Bedrock · global | $0.04/M cache read | input_cache_read: 0.00000004 / token (>272000; endpoint amazon-bedrock/global) | |
| OpenRouterOfficial partner | OpenAI: GPT-5.6 Luna ProPrompt > 272Kvia Azure | $0.04/M cache read | input_cache_read: 0.00000004 / token (>272000; endpoint azure) | |
| OpenRouterThird-party | PrismML: Ternary Bonsai 2 9B | $0.04/M cache read | input_cache_read: 0.00000004 / token (endpoint darkbloom/int4) | |
| OpenRouterThird-party | Qwen: Qwen3.5-9Bvia Darkbloom · fp4 | $0.04/M cache read | input_cache_read: 0.00000004 / token (endpoint darkbloom/fp4) | |
| OpenRouterOfficial | Qwen: Qwen3.7 FlashPrompt > 256Kvia Alibaba | $0.04/M cache read | input_cache_read: 0.00000004 / token (>256000; endpoint alibaba) | |
| OpenRouterThird-party | Qwen: Qwen3.8 27B (batch) | $0.04/M cache read | input_cache_read: 0.00000004 / token (endpoint deepinfra/bf16) | |
| OpenRouterThird-party | StepFun: Step 3.7 Flashvia Novita · fp8 | $0.04/M cache read | input_cache_read: 0.00000004 / token (endpoint novita/fp8) | |
| AibermNot disclosed | Claude Haiku 4.5 | $0.04/M cache read | model_ratio 0.875 × $2/M · group default ×0.23 · cache_ratio 0.1 | |
| AibermNot disclosed | Claude Haiku 4.5 Thinking | $0.04/M cache read | model_ratio 0.875 × $2/M · group default ×0.23 · cache_ratio 0.1 | |
| OpenRouterThird-party | MiniMax: MiniMax M2.7via GMICloud · fp8 | $0.042/M cache read | input_cache_read: 0.000000042 / token (endpoint gmicloud/fp8) | |
| OpenRouterThird-party | Tencent: Hy4 previewpeak 0000–1600 UTC · via Tencent · fp8 | $0.042/M cache read | input_cache_read: 0.000000042 / token (utc_start 0, utc_end 1600; endpoint tencent/fp8) | |
| OpenRouterThird-party | MiniMax: MiniMax M3 (batch) | $0.045/M cache read | input_cache_read: 0.0000000448 / token (endpoint deepinfra/fp8) | |
| AibermNot disclosed | Claude Sonnet 4.5 20250929 | $0.045/M cache read | model_ratio 0.9775 × $2/M · group default ×0.23 · cache_ratio 0.1 | |
| AibermNot disclosed | Claude Sonnet 4.5 20250929 Thinking | $0.045/M cache read | model_ratio 0.9775 × $2/M · group default ×0.23 · cache_ratio 0.1 | |
| AibermNot disclosed | Claude Sonnet 4.6 | $0.045/M cache read | model_ratio 0.9775 × $2/M · group default ×0.23 · cache_ratio 0.1 | |
| AibermNot disclosed | Claude Sonnet 4.6 Thinking | $0.045/M cache read | model_ratio 0.9775 × $2/M · group default ×0.23 · cache_ratio 0.1 | |
| RunwareNot disclosed | Shieldstral 1.0 3BPer 1M | $0.045/M cache read | $4.5e-8 / cachedInputToken | |
| AibermNot disclosed | Gemini 4 Argon | $0.046/M cache read | model_ratio 1 × $2/M · group default ×0.23 · cache_ratio 0.1 | |
| AibermNot disclosed | GPT-5.6 Terra | $0.049/M cache read | model_ratio 0.7065 × $2/M · group default ×0.23 · cache_ratio 0.15 | |
| AibermNot disclosed | Claude Opus 5 | $0.05/M cache read | model_ratio 1.086 × $2/M · group default ×0.23 · cache_ratio 0.1 | |
| OpenRouterOfficial | Anthropic: Claude Haiku 4.5 (batch) | $0.05/M cache read | input_cache_read: 0.00000005 / token (endpoint anthropic) | |
| OpenRouterOfficial partner | Anthropic: Claude Haiku 5.5Prompt > 100Kvia Google · global | $0.05/M cache read | input_cache_read: 0.00000005 / token (>100000; endpoint google-vertex/global) | |
| OpenRouterOfficial | Anthropic: Claude Sonnet 5.5 (batch) | $0.05/M cache read | input_cache_read: 0.00000005 / token (endpoint anthropic) | |
| OpenRouterOfficial | Google: Gemini 3 Flash Previewvia Google AI Studio | $0.05/M cache read | input_cache_read: 0.00000005 / token (endpoint google-ai-studio) | |
| OpenRouterOfficial | Google: Gemini 3.1 Flash Litevia Google AI Studio | $0.05/M cache read | input_audio_cache: 0.00000005 / token (endpoint google-ai-studio) | |
| OpenRouterOfficial | Google: Gemini 3.1 Flash Lite Previewvia Google AI Studio | $0.05/M cache read | input_audio_cache: 0.00000005 / token (endpoint google-ai-studio) | |
| OpenRouterOfficial | Google: Gemini 4 Argon (batch) | $0.05/M cache read | input_audio_cache: 0.00000005 / token (endpoint google-vertex/global) | |
| OpenRouterOfficial | Google: Gemini 4 Argon (batch) | $0.05/M cache read | input_cache_read: 0.00000005 / token (endpoint google-vertex/global) | |
| OpenRouterOfficial | Google: Gemini 4 Carbon (batch) | $0.05/M cache read | input_audio_cache: 0.00000005 / token (endpoint google-vertex/global) | |
| OpenRouterOfficial | Google: Gemini 4 Carbon (batch) | $0.05/M cache read | input_cache_read: 0.00000005 / token (endpoint google-vertex/global) | |
| OpenRouterThird-party | Google: Gemma 3 12Bvia NextBit · int4 | $0.05/M cache read | input_cache_read: 0.00000005 / token (endpoint nextbit/int4) | |
| OpenRouterOfficial | Inference.net: Schematron V2 Smallvia InferenceNet · bf16 | $0.05/M cache read | input_cache_read: 0.00000005 / token (endpoint inference-net/bf16) |
