DATUM

What does each model cost?

Every rate we hold, with the source named on each row. Six labs publish a pricing page we can read directly; where one does not, the number comes from a licensed index or from two independent registries — and where those two disagree, or only one has the model at all, the row is marked * rather than withheld. Looking for what a fixed level of capability costs rather than what a model costs? That is the AI deflation curve.

555 of 828 rates, 53 marked * — a single registry's word, or two that disagree. Those are the best available numbers rather than confirmed ones.

Sellerall 828OpenAI 122Together 87DeepInfra 68Google 42Z.ai 28Cloudflare 27Anthropic 26xAI 25Voyage 24Baseten 23Fireworks 22Alibaba 17Mistral 11Groq 10Amazon 6Meta-llama 5Moonshot 4AI21 2Cerebras 2Cohere 2DeepSeek 2
Ordercheapest output firstcheapest input firstby name· every tier and modality
ModelSellerIn /MtokOut /MtokCached inBeyond tokensNamed change
allam-2-7b*Groq$0$0——
GLM-4.5-FlashZ.ai$0$0$0—
GLM-4.6V-FlashZ.ai$0$0$0—
GLM-4.7-FlashZ.ai$0$0$0—
Ternary Bonsai 27BTogether$0$0——
GLM-OCRZ.ai$0.03$0.03——
Llama Guard 3 8BCloudflare$0.48$0.03——
Llama Prompt Guard 2 22MGroq$0.03$0.03——
Prompt Guard 2 86M*Groq$0.04$0.04——
Llama 3.1 8B*Groq$0.05$0.08——
gemma-3-4b-itDeepInfra$0.05$0.10——
gemma-4-E4B-itDeepInfra$0.02$0.10——
GLM-4-32B-0414-128KZ.ai$0.10$0.10——
ministral-3-3bMistral$0.10$0.10——
Granite 4.0 H MicroCloudflare$0.02$0.11——
qwen3.7-flashAlibaba$0.03$0.13$0.00—
GPT OSS 20BDeepInfra$0.03$0.14——
Llama 3 8B Instruct LiteTogether$0.14$0.14——
nova-microAmazon$0.04$0.14——
command-r7bCohere$0.04$0.15——
gemma-3-12b-itDeepInfra$0.05$0.15——
ministral-3-8bMistral$0.15$0.15——
Qwen3.5-9BDeepInfra$0.10$0.15——
Rnj-1 InstructTogether$0.15$0.15——
gemma-3-27b-itDeepInfra$0.08$0.16——
GPT OSS 120BDeepInfra$0.04$0.17——
DeepSeek-V4-FlashDeepInfra$0.09$0.18$0.02—
DeepSeek-V4-Flash-0731*DeepInfra$0.06$0.18$0.01—
llama-3.1-8bMeta-llama$0.18$0.18——
ministral-3-14bMistral$0.20$0.20——
Nemotron-3-Nano-30B-A3BDeepInfra$0.05$0.20$0.03—
nemotron-lightning-3p5-30b-a3bFireworks$0.05$0.20$0.01—
qwen-turboAlibaba$0.05$0.20——
Llama 3.2 1B InstructCloudflare$0.03$0.20——
nova-liteAmazon$0.06$0.24——
Qwen3.5 9BTogether$0.17$0.25——
DeepSeek-V4-Flash-0731*Baseten$0.13$0.26$0.03—
DeepSeek V4 Flash 0731Together$0.14$0.28——
MiMo-V2.5*DeepInfra$0.14$0.28$0.00—
Qwen3-32BDeepInfra$0.08$0.28——
Llama 3.1 8B Instruct fp8Cloudflare$0.15$0.29——
devstral-small-2Mistral$0.10$0.30——
gemma-4-26b-a4b-itCloudflare$0.10$0.30——
GPT OSS 20BGroq$0.07$0.30$0.04—
gpt-oss-20bCloudflare$0.20$0.30——
Llama 4 Scout 17B*DeepInfra$0.10$0.30——
Qwen2.5 7B Instruct TurboTogether$0.30$0.30——
Safety GPT OSS 20B*Groq$0.07$0.30——
Llama 3.3 70B Turbo*DeepInfra$0.10$0.32——
Llama 3.2 3B InstructCloudflare$0.05$0.34——
Qwen3 30B A3b fp8Cloudflare$0.05$0.34——
gemma-4-26B-A4B-itDeepInfra$0.07$0.34——
DeepSeek-V3.2DeepInfra$0.26$0.38$0.13—
gemma-4-31B-itDeepInfra$0.13$0.38——
Qwen3.8-Flash*DeepInfra$0.11$0.38$0.01—
babbage-002longOpenAI$0.40$0.40——
babbage-002shortOpenAI$0.40$0.40——
Gemini 2.5 Flash-LiteGoogle$0.10$0.40——
GLM-4.6V-FlashXZ.ai$0.04$0.40$0.00—
GLM-4.7-FlashDeepInfra$0.06$0.40$0.01—
GLM-4.7-FlashCloudflare$0.06$0.40——
GLM-4.7-FlashXZ.ai$0.07$0.40$0.01—
gpt-4.1-nanoshortOpenAI$0.10$0.40$0.03—
gpt-4.1-nanolongOpenAI$0.10$0.40$0.03—
gpt-5-nanolongOpenAI$0.05$0.40$0.01—
gpt-5-nanoshortOpenAI$0.05$0.40$0.01—
jamba-miniAI21$0.20$0.40——
Llama-3.3-Nemotron-Super-49B-v1.5*DeepInfra$0.40$0.40——
qwen-flashAlibaba$0.05$0.40$0.01—
qwen3.5-flashAlibaba$0.10$0.40$0.01—
Seed-2.0-miniDeepInfra$0.10$0.40$0.02—
Qwen3.8 FlashTogether$0.15$0.47——
qwen3.8-flashAlibaba$0.15$0.47——
Glm 5.3 Flash*Cloudflare$0.15$0.50$0.03—
GLM 5.3 Flash*Baseten$0.15$0.50——
GLM 5.3 Flash*Fireworks$0.15$0.50$0.03—
GLM-5.3-FlashTogether$0.15$0.50——
GLM-5.3-FlashZ.ai$0.15$0.50$0.03—
GLM-5.3-Flash*DeepInfra$0.15$0.50$0.03—
OpenAI GPT 120B*Baseten$0.10$0.50——
Qwen3-30B-A3BDeepInfra$0.12$0.50——
Qwen3-235B-A22B-Instruct-2507DeepInfra$0.09$0.55——
Gemma Sea Lion V4 27B ItCloudflare$0.35$0.56——
Mistral Small 3.1 24B InstructCloudflare$0.35$0.56——
Hy3DeepInfra$0.14$0.58$0.04—
llama-4-scoutMeta-llama$0.18$0.59——
command-rCohere$0.15$0.60——
deepseek-flashDeepSeek$0.15$0.60$0.00—
DeepSeek-V4.1-Flash*DeepInfra$0.20$0.60$0.01—
GPT OSS 120BFireworks$0.15$0.60$0.01—
gpt-4o-minishortOpenAI$0.15$0.60$0.07—
gpt-4o-minilongOpenAI$0.15$0.60$0.07—
gpt-oss-120bGroq$0.15$0.60$0.07—
gpt-oss-120BTogether$0.15$0.60——
mistral-small-4Mistral$0.15$0.60——
Qwen3 235B A22B Instruct 2507 FP8 ThroughputTogether$0.20$0.60——
deepseek-v4-flash-0731Fireworks$0.22$0.66$0.01—
deepseek-v4-flash-vision-expFireworks$0.22$0.66$0.01—
deepseek-v4p1-flashFireworks$0.22$0.66$0.01—
Llama 3.2 11B Vision InstructCloudflare$0.05$0.68——
gpt-oss-120bCerebras$0.35$0.75——
gpt-oss-120bCloudflare$0.35$0.75——
Nemotron Super*Baseten$0.30$0.75$0.06—
Llama 3.3 70B*Groq$0.59$0.79——
gpt-4.1-nano-2025-04-14OpenAI$0.20$0.80$0.05—
Llama 4 Maverick 17B FP8*DeepInfra$0.20$0.80——
Nemotron-3-Nano-Omni-30B-A3B-Reasoning*DeepInfra$0.20$0.80——
Llama 4 Scout 17B 16E InstructCloudflare$0.27$0.85——
llama-4-maverickMeta-llama$0.27$0.85——
Qwen3-VL-235B-A22B-InstructDeepInfra$0.20$0.88$0.11—
DeepSeek-V3DeepInfra$0.32$0.89——
codestral-25.08Mistral$0.30$0.90——
DeepSeek-V3-0324DeepInfra$0.24$0.90$0.14—
GLM-4.6VZ.ai$0.30$0.90$0.05—
DeepSeek-V3.1DeepInfra$0.25$0.95$0.13—
Qwen3.6 35B A3BDeepInfra$0.10$0.95——
Gemma 4 31BTogether$0.39$0.97——
MiniMax-M2.7DeepInfra$0.25$1$0.05—
Qwen 3.5 35B A3B*DeepInfra$0.14$1$0.05—
Qwen2.5 Coder 32B InstructCloudflare$0.66$1——
Qwen3 Coder 480B A35B Instruct TurboDeepInfra$0.30$1$0.10—
Qwq 32BCloudflare$0.66$1——
Llama 3.3 70BTogether$1.04$1.04——
llama-3.3-70bMeta-llama$1.04$1.04——
GLM-4.5-AirZ.ai$0.20$1.10$0.03—
MiniMax-M3DeepInfra$0.28$1.10$0.06—
Qwen3-Next-80B-A3B-InstructDeepInfra$0.09$1.10——
MiniMax M2.5*DeepInfra$0.15$1.15$0.03—
Step-3.7-FlashDeepInfra$0.20$1.15$0.04—
DeepSeek V4.1 FlashTogether$0.30$1.20——
DeepSeek-V4.1-Flash*Baseten$0.30$1.20$0.03—
gpt-4o-mini-2024-07-18OpenAI$0.30$1.20$0.15—
gpt-5.6-lunashortOpenAI$0.20$1.20$0.02—
inkling-small*Baseten$0.50$1.20$0.10—
Inkling-SmallDeepInfra$0.45$1.20$0.10—
MiniMax M2.7Together$0.30$1.20——
MiniMax M3Together$0.30$1.20——
MiniMax-M2.5Baseten$0.30$1.20——
MiniMax-M3Fireworks$0.30$1.20$0.06—
qwen-plusAlibaba$0.40$1.20$0.04—
Cogito v2.1 671BTogether$1.25$1.25——
gpt-5.4-nanolongOpenAI$0.20$1.25$0.02—
gpt-5.4-nanoshortOpenAI$0.20$1.25$0.02—
Qwen3.7-PlusTogether$0.32$1.28——
deepseek-v4-flash-0731*Cloudflare$0.44$1.32$0.01—
DeepSeek-V4-Flash-Vision-Exp*DeepInfra$0.44$1.32$0.01—
qwen-3.8-27bCerebras$0.99$1.49——
DeepSeek V3.1Baseten$0.50$1.50——
Gemini 3.1 Flash Lite Image 🍌Google$0.25$1.50——
Gemini 3.1 Flash-LiteGoogle$0.25$1.50——
gpt-3.5-turbolongOpenAI$0.50$1.50——
gpt-3.5-turboshortOpenAI$0.50$1.50——
gpt-3.5-turbo-0125shortOpenAI$0.50$1.50——
gpt-3.5-turbo-0125longOpenAI$0.50$1.50——
magistral-smallMistral$0.50$1.50——
mistral-large-3Mistral$0.50$1.50——
Muse Glimmer 30BTogether$0.35$1.50——
muse-glimmer-30bFireworks$0.35$1.50$0.04—
Nemotron 3 Super 120B*Cloudflare$0.50$1.50——
qwen3-coder-flashAlibaba$0.30$1.50$0.03—
qwen3-coder-nextAlibaba$0.30$1.50——
qwen3.6-flashAlibaba$0.25$1.50$0.03—
babbage-002OpenAI$1.60$1.60——
gpt-4.1-minishortOpenAI$0.40$1.60$0.10—
gpt-4.1-minilongOpenAI$0.40$1.60$0.10—
Qwen 3.7 Plus*Fireworks$0.40$1.60$0.08—
qwen3.7-plusAlibaba$0.40$1.60$0.04—
GLM-4.7DeepInfra$0.40$1.75$0.08—
GLM-4.5VZ.ai$0.60$1.80$0.11—
gpt-5.6-lunalongOpenAI$0.40$1.80$0.04—
deepseek-v4-proDeepSeek$0.66$1.98$0.02—
davinci-002shortOpenAI$2$2——
davinci-002longOpenAI$2$2——
devstral-2Mistral$0.40$2——
Gemini 2.5 Flash Native AudioGoogle$0.50$2——
GLM-4.6DeepInfra$0.50$2$0.10—
gpt-3.5-turbo-1106longOpenAI$1$2——
gpt-3.5-turbo-1106shortOpenAI$1$2——
gpt-3.5-turbo-instructshortOpenAI$1.50$2——
gpt-3.5-turbo-instructlongOpenAI$1.50$2——
gpt-5-minilongOpenAI$0.25$2$0.03—
gpt-5-minishortOpenAI$0.25$2$0.03—
grok-build-0.1< 200k prompt tokensxAI$1$2$0.20—
GLM-5DeepInfra$0.60$2.08$0.12—
DeepSeek-R1-0528DeepInfra$0.50$2.15$0.35—
GLM 4.7Baseten$0.60$2.20$0.12—
GLM-4.5Z.ai$0.60$2.20$0.11—
GLM-4.6Z.ai$0.60$2.20$0.11—
GLM-4.7Z.ai$0.60$2.20$0.11—
Kimi K2.5DeepInfra$0.45$2.25$0.07—
Llama 3.3 70B Instruct fp8 FastCloudflare$0.29$2.25——
GLM-5.2DeepInfra$0.75$2.40$0.14—
gpt-audio-miniOpenAI$0.60$2.40——
gpt-realtime-2.1-miniOpenAI$0.60$2.40$0.06—
gpt-realtime-miniOpenAI$0.60$2.40$0.06—
Nemotron Ultra*Baseten$0.60$2.40$0.12—
nemotron-3-ultra-nvfp4Fireworks$0.60$2.40$0.12—
Qwen3.5-122B-A10BDeepInfra$0.29$2.40——
qwen3.5-plusAlibaba$0.40$2.40$0.04—
Gemini 2.5 FlashGoogle$0.30$2.50——
Gemini 3.5 Flash-LiteGoogle$0.30$2.50——
grok-4.20-0309-non-reasoning< 200k prompt tokensxAI$1.25$2.50$0.20—
grok-4.20-0309-reasoning< 200k prompt tokensxAI$1.25$2.50$0.20—
grok-4.20-multi-agent-0309< 200k prompt tokensxAI$1.25$2.50$0.20—
grok-4.3< 200k prompt tokensxAI$1.25$2.50$0.20—
nova-2-liteAmazon$0.30$2.50——
DeepSeek-V4-ProDeepInfra$1.30$2.60$0.10—
DeepSeek-V4-Pro-0813DeepInfra$1.30$2.60$0.10—
Qwen3.5-27BDeepInfra$0.26$2.60——
Gemini 3 Flash PreviewGoogle$0.50$3——
Gemini 3.1 Flash Image 🍌Google$0.50$3——
Kimi K2.5Baseten$0.60$3$0.12—
MiMo-V2.5-ProDeepInfra$1$3$0.20—
Qwen 3.5 397B A17B*DeepInfra$0.45$3$0.22—
qwen3.6-27bGroq$0.60$3$0.30—
qwen3.6-plusAlibaba$0.50$3$0.05—
Qwen3.6-PlusTogether$0.50$3——
Qwen3.8-27BDeepInfra$0.40$3$0.04—
Seed-2.0-codeDeepInfra$0.50$3$0.10—
Seed-2.0-proDeepInfra$0.50$3$0.10—
GLM 5Baseten$0.95$3.15$0.20—
GLM-5Z.ai$1$3.20$0.20—
gpt-4.1-mini-2025-04-14OpenAI$0.80$3.20$0.20—
nova-proAmazon$0.80$3.20——
Qwen3.6-27BDeepInfra$0.32$3.20——
qwen3.8-27b*Cloudflare$0.45$3.20$0.05—
Kimi-K2.7-CodeDeepInfra$0.68$3.40$0.14—
DeepSeek-V4-Pro*Baseten$1.74$3.48$0.14—
GLM-5.1DeepInfra$1.05$3.50$0.20—
Kimi K2.6DeepInfra$0.75$3.50$0.15—
llama-3.1-405bMeta-llama$3.50$3.50——
Qwen3.5-397B-A17BTogether$0.60$3.60——
Gemini 3.6 FlashGoogle$0.75$3.75——$7.50 from Fri Jan 01 2027 00:00:00 GMT+0000 (Coordinated Universal Time)
Gemini 3.7 FlashGoogle$0.75$3.75——$7.50 from Fri Jan 01 2027 00:00:00 GMT+0000 (Coordinated Universal Time)
Gemini 3.8 FlashGoogle$0.75$3.75——$7.50 from Fri Jan 01 2027 00:00:00 GMT+0000 (Coordinated Universal Time)
DeepSeek V4 Pro 0813Together$1.32$3.96——
deepseek-v4-pro-0813Fireworks$1.32$3.96$0.04—
deepseek-v4-pro-0813*Cloudflare$1.32$3.96$0.04—
DeepSeek-V4-Pro-0813*Baseten$1.32$3.96——
Claude Haiku 3.5Anthropic$0.80$4$0.08—
GLM-5.3*DeepInfra$1.20$4$0.12—
grok-build-0.1≥ 200k prompt tokensxAI$2$4$0.40—
Kimi K2.6*Baseten$0.95$4$0.16—
Kimi K2.6*Fireworks$0.95$4$0.16—
Kimi K2.6Moonshot$0.95$4$0.16—
Kimi K2.7 Code*Fireworks$0.95$4$0.19—
kimi-k2.6Cloudflare$0.95$4$0.16—
kimi-k2.7-codeCloudflare$0.95$4$0.19—
kimi-k2.7-codeMoonshot$0.95$4$0.19—
Kimi-K2.7-Code*Baseten$0.95$4$0.16—
qwen3.8-27bGroq$0.80$4——
inkling*Baseten$1$4.05——
inklingFireworks$1$4.05$0.17—
InklingTogether$1$4.05——
InklingDeepInfra$0.95$4.05$0.16—
GLM 5.1*Baseten$1.30$4.30$0.26—
Glm 5.2Cloudflare$1.40$4.40$0.26—
GLM 5.2*Fireworks$1.40$4.40$0.14—
GLM 5.2*Baseten$1.40$4.40$0.30—
Glm 5.3*Cloudflare$1.40$4.40$0.26—
GLM 5.3Baseten$1.40$4.40$0.14—
GLM 5.3*Fireworks$1.40$4.40$0.26—
GLM-5.1Z.ai$1.40$4.40$0.26—
GLM-5.2Together$1.40$4.40——
GLM-5.2Z.ai$1.40$4.40$0.26—
GLM-5.3Together$1.40$4.40——
GLM-5.3Z.ai$1.40$4.40$0.26—
o1-miniOpenAI$1.10$4.40$0.55—
o3-minishortOpenAI$1.10$4.40$0.55—
o3-minilongOpenAI$1.10$4.40$0.55—
o4-minishortOpenAI$1.10$4.40$0.28—
o4-minilongOpenAI$1.10$4.40$0.28—
Gemini 3.1 Flash Live PreviewGoogle$0.75$4.50——
GLM-4.5-AirXZ.ai$1.10$4.50$0.22—
gpt-5.4-minilongOpenAI$0.75$4.50$0.07—
gpt-5.4-minishortOpenAI$0.75$4.50$0.07—
Deepseek R1 Distill Qwen 32BCloudflare$0.50$4.88——
Qwen3.8-MaxDeepInfra$1.65$4.95$0.21—
Claude Haiku 4.5Anthropic$1$5$0.10—
claude-haiku-4-5-20251001Anthropic$1$5$0.10—
Gemini Robotics ER 2 PreviewGoogle$1$5——$10 from Fri Jan 01 2027 00:00:00 GMT+0000 (Coordinated Universal Time)
gpt-4o-mini-transcribeOpenAI$1.25$5——
grok-4.20-0309-non-reasoning≥ 200k prompt tokensxAI$2.50$5$0.40—
grok-4.20-0309-reasoning≥ 200k prompt tokensxAI$2.50$5$0.40—
grok-4.20-multi-agent-0309≥ 200k prompt tokensxAI$2.50$5$0.40—
grok-4.3≥ 200k prompt tokensxAI$2.50$5$0.40—
magistral-mediumMistral$2$5——
qwen3-coder-plusAlibaba$1$5$0.10—
gpt-3.5-turboOpenAI$3$6——
grok-4.5< 200k prompt tokensxAI$2$6$0.30—
grok-4.6< 200k prompt tokensxAI$2$6$0.50—
qwen3-maxAlibaba$1.20$6$0.12—
Qwen3-MaxDeepInfra$1.20$6$0.24—
Qwen3.7-MaxTogether$2$6——
Qwen3.8-2.4T-A95BTogether$2$6——
Qwen3.8-2.4T-A95BDeepInfra$2$6$0.20—
qwen3.8-maxAlibaba$2$6——
qwen3p8-2p4t-a95b*Fireworks$2$6$0.25—
qwen3p8-maxFireworks$2$6$0.25—
qwen-maxAlibaba$1.60$6.40——
GLM 5.2 Fast*Baseten$2.10$6.60$0.21—
GLM 5.2 Fast*Fireworks$2.10$6.60$0.21—
GLM 5.3 Fast*Fireworks$2.10$6.60$0.39—
GLM 5.3 Fast*Baseten$2.10$6.60——
mistral-medium-3.5Mistral$1.50$7.50——
qwen3.7-maxAlibaba$2.50$7.50$0.25—
Qwen3.7-MaxDeepInfra$2.50$7.50$0.50—
gpt-4.1shortOpenAI$2$8$0.50—
gpt-4.1longOpenAI$2$8$0.50—
jamba-largeAI21$2$8——
Kimi K2.7 Code HighSpeed*Moonshot$1.90$8$0.38—
o3shortOpenAI$2$8$0.50—
o3longOpenAI$2$8$0.50—
o4-mini-deep-researchOpenAI$2$8$0.50—
GLM-4.5-XZ.ai$2.20$8.90$0.45—
Gemini 3.5 FlashGoogle$1.50$9——
Gemini Omni FlashGoogle$1.50$9——
Gemini Omni Flash PreviewGoogle$1.50$9——
chatgpt-image-latestOpenAI$5$10$1.25—
Claude Sonnet 5Anthropic$2$10$0.20—
Gemini 2.5 Computer Use PreviewGoogle$1.25$10——
Gemini 2.5 Flash Preview TTSGoogle$0.50$10——
Gemini 2.5 ProGoogle$1.25$10——
Gemini Robotics ER 2 Streaming PreviewGoogle$1.25$10——$10 from Fri Jan 01 2027 00:00:00 GMT+0000 (Coordinated Universal Time)
gpt-4oshortOpenAI$2.50$10$1.25—
gpt-4olongOpenAI$2.50$10$1.25—
gpt-4o-transcribeOpenAI$2.50$10——
gpt-4o-transcribe-diarizeOpenAI$2.50$10——
gpt-5shortOpenAI$1.25$10$0.13—
gpt-5longOpenAI$1.25$10$0.13—
gpt-5-search-apiOpenAI$1.25$10$0.13—
gpt-5.1longOpenAI$1.25$10$0.13—
gpt-5.1shortOpenAI$1.25$10$0.13—
gpt-audioOpenAI$2.50$10——
gpt-audio-1.5OpenAI$2.50$10——
gpt-image-1.5OpenAI$5$10$1.25—
nova-2-proAmazon$1.25$10——
davinci-002OpenAI$12$12——
Gemini 3 Pro Image 🍌Google$2$12——
Gemini 3.1 Pro PreviewGoogle$2$12——
Gemini 3.5 TranscribeGoogle$2$12——
gpt-4.1-2025-04-14OpenAI$3$12$0.75—
gpt-5.6-terrashortOpenAI$2$12$0.20—
grok-4.5≥ 200k prompt tokensxAI$4$12$0.60—
grok-4.6≥ 200k prompt tokensxAI$4$12$1—
nova-premierAmazon$2.50$13——
gpt-5.2shortOpenAI$1.75$14$0.17—
gpt-5.2longOpenAI$1.75$14$0.17—
gpt-5.3-codexOpenAI$1.75$14——
Kimi-K3DeepInfra$2.85$14$0.28—
Claude Sonnet 4Anthropic$3$15$0.30—
Claude Sonnet 4.5Anthropic$3$15$0.30—
Claude Sonnet 4.6Anthropic$3$15$0.30—
claude-sonnet-4-20250514Anthropic$3$15$0.30—
claude-sonnet-4-5-20250929Anthropic$3$15$0.30—
gpt-4o-2024-05-13shortOpenAI$5$15——
gpt-4o-2024-05-13longOpenAI$5$15——
gpt-4o-2024-08-06OpenAI$3.75$15$1.88—
gpt-5.4shortOpenAI$2.50$15$0.25—
Kimi K3Together$3$15——
kimi-k3Fireworks$3$15$0.30—
kimi-k3Moonshot$3$15$0.30—
Kimi-K3*Baseten$3$15——
gpt-realtimeOpenAI$4$16$0.40—
gpt-realtime-1.5OpenAI$4$16$0.40—
gpt-5.6-terralongOpenAI$4$18$0.40—
Gemini 2.5 Pro Preview TTSGoogle$1$20——
Gemini 3.1 Flash TTS PreviewGoogle$1$20——
gpt-5.6-solshortOpenAI$4$20$0.40—
Gemini 3.5 Live TranslateGoogle$3.50$21——
Gemini 3.5 Transcribe LiveGoogle$3.50$21——
gpt-5.4longOpenAI$5$23$0.50—
Kimi K3 FastFireworks$4.50$23$0.45—
gpt-realtime-2OpenAI$4$24$0.40—
gpt-realtime-2.1OpenAI$4$24$0.40—
Claude Opus 4.5Anthropic$5$25$0.50—
Claude Opus 4.6Anthropic$5$25$0.50—
Claude Opus 4.7Anthropic$5$25$0.50—
Claude Opus 4.8Anthropic$5$25$0.50—
Claude Opus 5Anthropic$5$25$0.50—
claude-opus-4-5-20251101Anthropic$5$25$0.50—
gpt-rosalind-researchOpenAI$5$25$0.50—
chat-latestOpenAI$5$30$0.50—
gpt-4-turbo-2024-04-09longOpenAI$10$30——
gpt-4-turbo-2024-04-09shortOpenAI$10$30——
gpt-5.5shortOpenAI$5$30$0.50—
gpt-5.6-sollongOpenAI$8$30$0.80—
o3-deep-researchOpenAI$10$40$2.50—
gpt-5.5longOpenAI$10$45$1—
Claude Fable 5Anthropic$10$50$1—
Claude Fable 5.1Anthropic$10$50$0.25—
Claude Mythos 5Anthropic$10$50$1—
Claude Mythos 5.1Anthropic$10$50$0.25—
gpt-6-astrashortOpenAI$10$50$1—
gpt-4-0613longOpenAI$30$60——
gpt-4-0613shortOpenAI$30$60——
o1longOpenAI$15$60$7.50—
o1shortOpenAI$15$60$7.50—
Claude Opus 4Anthropic$15$75$1.50—
Claude Opus 4.1Anthropic$15$75$1.50—
claude-opus-4-1-20250805Anthropic$15$75$1.50—
claude-opus-4-20250514Anthropic$15$75$1.50—
gpt-5.5-cybershortOpenAI$13$75$1.25—
gpt-5.5-cyberlongOpenAI$13$75$1.25—
gpt-5.6-cybershortOpenAI$13$75$1.25—
gpt-5.6-cyberlongOpenAI$13$75$1.25—
gpt-6-astralongOpenAI$20$75$2—
o3-proshortOpenAI$20$80——
o3-prolongOpenAI$20$80——
gpt-5-proshortOpenAI$15$120——
gpt-5-prolongOpenAI$15$120——
gpt-5.2-proshortOpenAI$21$168——
gpt-5.2-prolongOpenAI$21$168——
gpt-5.4-proshortOpenAI$30$180——
gpt-5.5-proshortOpenAI$30$180——
gpt-5.4-prolongOpenAI$60$270——
gpt-5.5-prolongOpenAI$60$270——
o1-proshortOpenAI$150$600——
o1-prolongOpenAI$150$600——
1000px x 1000pxPrice per thousand imagesVoyage———$0.60 per thousand images
1000px x 1000pxPrice per imageVoyage———$0.00 per image
2000px x 2000pxPrice per imageVoyage———$0.00 per image
2000px x 2000pxPrice per thousand imagesVoyage———$1.20 per thousand images
200px x 200pxPrice per imageVoyage———$0.00 per image
200px x 200pxPrice per thousand imagesVoyage———$0.03 per thousand images
4000px x 4000pxPrice per imageVoyage———$0.00 per image
4000px x 4000pxPrice per thousand imagesVoyage———$1.20 per thousand images
Agent KitPricingOpenAI$0———
ByteDance Seedance 1.0 LiteTogether———$0.14 per video
ByteDance Seedance 1.0 ProTogether———$0.57 per video
ByteDance Seedance 2.0Together———$0.16 per video
ByteDance Seedance 2.5Together———$0.12 per video
ByteDance Seedream 3.0Together$0.02———
ByteDance Seedream 4.0Together$0.03———
Cartesia Sonic-2Together———$65 per million characters
Cartesia Sonic-3Together———$65 per million characters
CCU priceDetailsAnthropic———$0.01 per hour
Code executionFree TierGoogle$0———
Code ExecutionCost / 1k CallsxAI———$5 per thousand calls
CogVideoX-3Z.ai———$0.20 per video†
CogView-4Z.ai———$0.01 per image†
Collections SearchCost / 1k CallsxAI———$2.50 per thousand calls
Context cachingFree TierGoogle$0———
Context caching priceFree TierGoogle$0———
ConversionDetailsAnthropic———$0.01 per hour
File AttachmentsCost / 1k CallsxAI———$10 per thousand calls
File searchPaid Tier, per 1M tokens in USDGoogle$0.15———
FLUX 3Together———$0.17 per video
FLUX.1 Kontext [max]Together$0.08———
FLUX.1 Kontext [pro]Together$0.04———
FLUX.2 [max]Together$0.07———
FLUX1.1 [pro]Together$0.04———
Gemini 2.5 Flash Image 🍌Google$0.30———
Gemini 2.5 Flash Image 🍌Google———$0.04 per image
Gemini Embedding 2Google$12———
General-Purpose TranslationZ.ai$3———
GLM Slide/Poster AgentZ.ai$0.70———
GLM-ASR-2512Z.ai$0.03———
GLM-ImageZ.ai———$0.01 per image†
Google Imagen 4.0 FastTogether$0.02———
Google Imagen 4.0 PreviewTogether$0.04———
Google Imagen 4.0 UltraTogether$0.06———
Google MapsPaid Tier, per 1M tokens in USDGoogle$0———
Google SearchFree TierGoogle$0———
Google Veo 2.0Together———$2.50 per video
gpt-4o-mini-ttsOpenAI$0.60———
gpt-image-1OpenAI$5—$1.25—
gpt-image-1-miniOpenAI$2—$0.20—
gpt-image-2OpenAI$5—$1.25—
gpt-image-2.5-flareOpenAI$5—$1.25—
gpt-image-2.5-sunburstOpenAI$5—$1.25—
gpt-live-1OpenAI———$0.05 per minute
grok-imagine-imagexAI———$0.02 per image
grok-imagine-image-2.0xAI———$0.04 per image
grok-imagine-image-qualityxAI———$0.05 per image
grok-imagine-videoxAI———$0.05 per second
grok-imagine-video-1.5xAI———$0.08 per second
HappyHorse 1.0 I2VTogether———$0.24 per video
HappyHorse 1.0 R2VTogether———$0.24 per video
HappyHorse 1.0 T2VTogether———$0.24 per video
HappyHorse 1.1 I2VTogether———$0.14 per video
HappyHorse 1.1 R2VTogether———$0.14 per video
HappyHorse 1.1 T2VTogether———$0.14 per video
Ideogram 3.0Together$0.06———
InklingTogether———$1 per million characters
Input priceFree TierGoogle$0———
Juggernaut Lightning FluxTogether$0.00———
Juggernaut Pro FluxTogether$0.00———
Kling 1.6 StandardTogether———$0.19 per video
Kling 2.1 MasterTogether———$0.92 per video
Kling 2.1 ProTogether———$0.32 per video
Kling 2.1 StandardTogether———$0.18 per video
Kokoro-82M TTSTogether———$4 per million characters
Lyria 3 Clip PreviewPaid Tier, per request in USDGoogle———$0.04 per request
Lyria 3 Pro PreviewPaid Tier, per request in USDGoogle———$0.08 per request
Lyria 3.5Paid Tier, per request in USDGoogle———$0.08 per request
MiniMax 01 DirectorTogether———$0.28 per video
MiniMax Hailuo 02Together———$0.49 per video
NVIDIA HGX B200Preemptible ComputeTogether———$4.09 per hour
NVIDIA HGX B200ReservedTogether———$7.99 per hour
NVIDIA HGX B200HourlyTogether———$8.19 per hour
NVIDIA HGX B200On-demand Pay as you goTogether———$8.99 per hour
NVIDIA HGX B200ON-DemandTogether———$8.19 per hour
NVIDIA HGX H100HourlyTogether———$3.99 per hour
NVIDIA HGX H100Preemptible ComputeTogether———$1.99 per hour
NVIDIA HGX H100ON-DemandTogether———$3.99 per hour
NVIDIA HGX H100ReservedTogether———$3.69 per hour
NVIDIA HGX H200HourlyTogether———$5.99 per hour
NVIDIA HGX H200Preemptible ComputeTogether———$2.99 per hour
NVIDIA HGX H200ON-DemandTogether———$5.99 per hour
NVIDIA HGX H200ReservedTogether———$4.99 per hour
NVIDIA Nemotron 3 ASR Streaming 0.6BTogether$0.00———
NVIDIA Nemotron 3.5 ASRTogether$0.00———
NVIDIA Parakeet TDT 0.6B v3Together$0.00———
NVIDIA Parakeet TDT 0.6B V3 RealtimeTogether———$0.00 per million characters
omni-moderation-latestOpenAI$0———
Orpheus TTSTogether———$15 per million characters
Output priceFree TierGoogle$0———
Per GiB RAMPrice/HourTogether———$0.01 per hour
Per vCPUPrice/HourTogether———$0.04 per hour
PixVerse v5Together———$0.30 per video
Popular Special Effects Video TemplatesZ.ai———$0.20 per video
Qwen ImageTogether$0.01———
Qwen3.6-PlusTogether———$0.50 per video
rerank-2.5Voyage$0.05———
rerank-2.5Voyage———$0.00 per request
rerank-2.5-liteVoyage$0.02———
rerank-2.5-liteVoyage———$0.00 per request
rerank-3Voyage$0.05———
rerank-3Voyage———$0.00 per request
rerank-3-liteVoyage———$0.00 per request
rerank-3-liteVoyage$0.02———
SD XLTogether$0.00———
Session runtimeRateAnthropic———$0.08 per session
Sora 2Together———$0.80 per video
sora-2OpenAI———$0.10 per second
Text to SpeechxAI———$15 per million characters
text-embedding-3-largeOpenAI$0.13———
text-embedding-3-smallOpenAI$0.02———
text-embedding-ada-002OpenAI$0.10———
tts-1OpenAI———$15 per million characters
tts-1-hdOpenAI———$30 per million characters
URL contextFree TierGoogle$0———
Vidu Q1Together———$0.22 per video
voyage-4Voyage$0.06———
voyage-4-largeVoyage$0.12———
voyage-4-liteVoyage$0.02———
voyage-code-4Voyage$0.12———
voyage-context-4Voyage$0.12———
voyage-finance-2 voyage-law-2 voyage-code-2Voyage$0.12———
voyage-multimodal-3.5 voyage-multimodal-3Voyage———$0.60 per billion pixels
voyage-multimodal-3.5 voyage-multimodal-3Voyage$0.12———
Web SearchZ.ai———$0.01 per use
Web SearchCost / 1k CallsxAI———$5 per thousand calls
X SearchCost / 1k CallsxAI———$5 per thousand calls
Beyond tokens99 rates in 12 units

Not everything is sold by the token. An image, a second of video, a minute of speech and a million characters are each their own measurement, and a table that answered only in tokens left them out. They are printed here in the unit the seller published them in, and never converted into one another.

per image25 rates

200px x 200pxVoyage$0.00per image
1000px x 1000pxVoyage$0.00per image
2000px x 2000pxVoyage$0.00per image
4000px x 4000pxVoyage$0.00per image
CogView-4Z.ai$0.01per image
GLM-ImageZ.ai$0.01per image
FLUX.2 [dev]Together$0.02per image
Gemini 2.5 Flash Image 🍌Google$0.02per image
Gemini 2.5 Flash Image 🍌Google$0.02per image
grok-imagine-imagexAI$0.02per image
FLUX.2 [flex]Together$0.03per image
FLUX.2 [pro]Together$0.03per image
Wan 2.6 ImageTogether$0.03per image
GPT Image 1.5Together$0.03per image

showing the cheapest 14 of 25

per video24 rates

ByteDance Seedance 2.5Together$0.12per video
ByteDance Seedance 1.0 LiteTogether$0.14per video
HappyHorse 1.1 I2VTogether$0.14per video
HappyHorse 1.1 R2VTogether$0.14per video
HappyHorse 1.1 T2VTogether$0.14per video
ByteDance Seedance 2.0Together$0.16per video
FLUX 3Together$0.17per video
Kling 2.1 StandardTogether$0.18per video
Kling 1.6 StandardTogether$0.19per video
CogVideoX-3Z.ai$0.20per video
Popular Special Effects Video TemplatesZ.ai$0.20per video
Vidu Q1Together$0.22per video
HappyHorse 1.0 I2VTogether$0.24per video
HappyHorse 1.0 R2VTogether$0.24per video

showing the cheapest 14 of 24

per hour17 rates

CCU priceAnthropic$0.01per hour
ConversionAnthropic$0.01per hour
Per GiB RAMTogether$0.01per hour
Per vCPUTogether$0.04per hour
NVIDIA HGX H100Together$1.99per hour
NVIDIA HGX H200Together$2.99per hour
NVIDIA HGX H100Together$3.69per hour
NVIDIA HGX H100Together$3.99per hour
NVIDIA HGX H100Together$3.99per hour
NVIDIA HGX B200Together$4.09per hour
NVIDIA HGX H200Together$4.99per hour
NVIDIA HGX H200Together$5.99per hour
NVIDIA HGX H200Together$5.99per hour
NVIDIA HGX B200Together$7.99per hour

showing the cheapest 14 of 17

per million characters9 rates

NVIDIA Parakeet TDT 0.6B V3 RealtimeTogether$0.00per million characters
InklingTogether$1per million characters
Kokoro-82M TTSTogether$4per million characters
Orpheus TTSTogether$15per million characters
Text to SpeechxAI$15per million characters
tts-1OpenAI$15per million characters
tts-1-hdOpenAI$30per million characters
Cartesia Sonic-2Together$65per million characters
Cartesia Sonic-3Together$65per million characters

per request7 rates

rerank-2.5-liteVoyage$0.00per request
rerank-3-liteVoyage$0.00per request
rerank-2.5Voyage$0.00per request
rerank-3Voyage$0.00per request
Lyria 3 Clip PreviewGoogle$0.04per request
Lyria 3 Pro PreviewGoogle$0.08per request
Lyria 3.5Google$0.08per request

per thousand calls5 rates

Collections SearchxAI$2.50per thousand calls
Code ExecutionxAI$5per thousand calls
Web SearchxAI$5per thousand calls
X SearchxAI$5per thousand calls
File AttachmentsxAI$10per thousand calls

per thousand images4 rates

200px x 200pxVoyage$0.03per thousand images
1000px x 1000pxVoyage$0.60per thousand images
2000px x 2000pxVoyage$1.20per thousand images
4000px x 4000pxVoyage$1.20per thousand images

per second4 rates

grok-imagine-videoxAI$0.05per second
sora-2OpenAI$0.05per second
grok-imagine-video-1.5xAI$0.08per second
sora-2OpenAI$0.10per second

per billion pixels1 rate

voyage-multimodal-3.5 voyage-multimodal-3Voyage$0.60per billion pixels

per minute1 rate

gpt-live-1OpenAI$0.05per minute

per session1 rate

Session runtimeAnthropic$0.08per session

per use1 rate

Web SearchZ.ai$0.01per use

† The seller names no unit on that table. It was read from the section it sits under, or from a column named for a token role, and it is marked rather than presented as the seller's own words.

What would it cost to actually do something?

A rate per million tokens is how the industry sells, and almost nobody can feel it. That question — one question, one long document, one day of an agent — has a page of its own, with the same numbers in the unit a person has.

How to read this table

Prices are per million tokens, in US dollars. Every row names the seller it came from, and where that is not the seller's own published page the mark says so: * means the number comes from a single registry or two that disagree, so it is the best available rather than a confirmed one. The same mark covers both because the remedy is the same — check the seller before you commit.

A row is a rate, not a model. One model can appear four times at four prices — standard, batch, flex, fast — and again for audio or image tokens, and again either side of a context threshold. This page opens on the standard text rates because those are the ones a reader compares; the toggle shows everything.

Where a seller has announced a rise, it is named here. A dated increase a reader can still act on is the most useful thing this table holds, so it sits beside the current rate rather than in a footnote.