AI Model API Prices on OpenRouter

We record the input and output prices per million tokens of 466 AI models from 60 providers as OpenRouter lists them. These are listed prices, not a billing test, and model quality is not rated. Prices as shown in 2 countries — the United States and Switzerland — on ; the rows mix countries, and each plan page names its own.

More on AI: Technical checks · Subscription limits · Terms tracker · What's new · GPU prices

How to cite this

Caliper Press, “AI API Price Watch”, $1.80 per 1M tokens, based on 345 models, as of Oct 18, 2026; methodology; licence CC BY 4.0.

Caliper Press, “AI API Price Watch — major-vendor flagships (direct vendor price)”, $6.00 per 1M tokens, based on 17 models, as of Oct 18, 2026; methodology; licence CC BY 4.0.

Caliper Press, “AI API Price Watch — major-vendor flagships (OpenRouter price)”, $5.06 per 1M tokens, based on 22 models, as of Oct 18, 2026; methodology; licence CC BY 4.0.

What typical tasks cost

Per-token prices are hard to compare, so we priced three typical tasks (1,000 chat replies, summarizing 100 documents and a day of coding assistant use) on 15 current models. At the providers' own prices, 1,000 chat replies cost from $0.25 (GPT-6 Luna) to $25.00 (Claude Fable 5.1 and GPT-6 Astra). Calculated LLM API pricing comparison by task

Not building on an API? Ready-made AI tools for writing, speech and images charge per month or per credit instead of per token: AI tool prices and the usage each plan includes.

Models by provider

Providers with at least 5 tracked models have their own page listing every model; models of the other providers are listed below the table.

ProviderModels trackedLowest paid output price per 1M tokens
OpenAI104$0.09 · OpenAI: gpt-oss-20b
Qwen55$0.13 · Qwen: Qwen3.7 Flash
Google44$0.10 · Google: Gemma 3 4B
Anthropic31$0.50 · Anthropic: Claude Haiku 5.5
Mistral26$0.03 · Mistral: Mistral Nemo
Z.ai18$0.20 · Z.ai: GLM 5.3 Flash
DeepSeek16$0.14 · DeepSeek: DeepSeek V4.1 Flash
Meta16$0.08 · Meta: Llama 3.1 8B Instruct
NVIDIA12$0.20 · NVIDIA: Nemotron 3.5 Lightning
MiniMax8$0.95 · MiniMax: MiniMax M2.5
MoonshotAI8$2.30 · MoonshotAI: Kimi K2 0711
SpaceXAI8$2.00 · SpaceXAI: Grok Build 0.1
Tencent7$0.18 · Tencent: Hy-MT2-1.8B
AionLabs6$1.40 · AionLabs: Aion 3.5 Mini
ByteDance Seed6$0.30 · ByteDance Seed: Seed 1.6 Flash
Cohere6$0.15 · Cohere: Command R7B (12-2024)
inclusionAI6$0.06 · inclusionAI: Ling 3.0 Flash
Amazon5$0.14 · Amazon: Nova Micro 1.0
Perplexity5$1.00 · Perplexity: Sonar
Xiaomi5$0.28 · Xiaomi: MiMo-V2.5
Poolside4$0.12 · Poolside: Laguna XS 2.1
Sakana4$4.00 · Sakana: Sakana Namazu
Thinking Machines4$1.20 · Thinking Machines: Inkling Small
Unbiased4$3.20 · Pareto 26.10
Nous3$0.70 · Nous: Hermes 3 70B Instruct
Sao10K3$0.05 · Sao10K: Llama 3 8B Lunaris
StepFun3$0.30 · StepFun: Step 3.5 Flash
TheDrummer3$0.40 · TheDrummer: UnslopNemo 12B
Upstage3$0.20 · Upstage: Solar Mini 4
IBM2$0.11 · IBM: Granite 4.0 Micro
Inception2$0.15 · Inception: Mercury 2.5
Inference.net2$0.15 · Inference.net: Schematron V2 Turbo
LiquidAI2$0.70 · LiquidAI: LFM2.5-24B-A2B
Microsoft2$0.14 · Microsoft: Phi 4
Moondream2$2.50 · Moondream: Moondream 3 Preview
Morph2$1.20 · Morph: Morph V3 Fast
Nex AGI2$0.10 · Nex AGI: Nex-N2.5-Mini
Perceptron2$1.50 · Perceptron: Perceptron Mk1
Rekaai2$0.10 · Reka Edge
Relace2$1.25 · Relace: Relace Apply 3
ScaleDown2Unknown
Anthracite-org1$5.00 · Magnum v4 72B
Apodex1Unknown
Arcee AI1$0.80 · Arcee AI: Trinity Large Thinking
Baidu1Unknown
ByteDance1$0.20 · ByteDance: UI-TARS 7B
Venice1$0.90 · Venice: Uncensored
Dots Studio1Unknown
Elorian1$3.00 · Elorian: Glimpse
Fireworks1$15.00 · Fireworks: Ember-1
Gryphe1$0.11 · MythoMax 13B
Kwaipilot1Unknown
Mancer1$0.75 · Mancer: Weaver (alpha)
Meituan1$1.20 · Meituan: LongCat 2.0
PrismML1$0.50 · PrismML: Ternary Bonsai 2 27B
Reflection1$3.00 · Reflection: Beam 1.0 501B A23B
Stealth1Unknown
Undi951$0.65 · ReMM SLERP 13B
Writer1$6.00 · Writer: Palmyra X5
XingChen1Unknown

Models of providers with fewer than 5 tracked models (74)

ProviderModelInput per 1M tokensOutput per 1M tokensContext (tokens)Listed since
scaledownScaleDown: Extract$0.04$0.001,000,000Oct 13, 2026
scaledownScaleDown: Summarize$0.04$0.001,000,000Oct 13, 2026
sao10kSao10K: Llama 3 8B Lunaris$0.04$0.058,192Aug 13, 2024
nex-agiNex AGI: Nex-N2.5-Mini$0.03$0.10262,144Sep 8, 2026
rekaaiReka Edge$0.10$0.1016,384Mar 20, 2026
grypheMythoMax 13B$0.08$0.118,192Jul 2, 2023
ibm-graniteIBM: Granite 4.0 Micro$0.02$0.11131,000Oct 20, 2025
poolsidePoolside: Laguna XS 2.1$0.06$0.12262,144Jul 2, 2026
microsoftMicrosoft: Phi 4$0.07$0.1416,384Jan 10, 2025
inference-netInference.net: Schematron V2 Turbo$0.03$0.15128,000Sep 12, 2026
inceptionInception: Mercury 2.5$0.04$0.15260,000Sep 8, 2026
poolsidePoolside: Laguna S 2.1$0.09$0.181,048,576Jul 21, 2026
upstageUpstage: Solar Mini 4$0.05$0.20524,288Sep 23, 2026
bytedanceByteDance: UI-TARS 7B $0.10$0.20128,000Jul 22, 2025
rekaaiReka Flash 3$0.10$0.2065,536Mar 12, 2025
inference-netInference.net: Schematron V2 Small$0.05$0.23128,000Sep 12, 2026
ibm-graniteIBM: Granite 4.2 8B$0.06$0.25131,072Aug 31, 2026
nex-agiNex AGI: Nex-N2.5-Pro$0.08$0.25262,144Sep 8, 2026
stepfunStepFun: Step 3.5 Flash$0.10$0.30262,144Jan 29, 2026
upstageUpstage: Solar Pro 4$0.09$0.36524,288Aug 10, 2026
thedrummerTheDrummer: UnslopNemo 12B$0.40$0.401,024,000Nov 8, 2024
prism-mlPrismML: Ternary Bonsai 2 27B$0.08$0.50262,144Sep 18, 2026
thedrummerTheDrummer: Cydonia 24B V4.1$0.30$0.50131,072Sep 27, 2025
upstageUpstage: Solar Pro 3$0.15$0.60131,072Jan 27, 2026
microsoftWizardLM-2 8x22B$0.62$0.6265,535Apr 16, 2024
undi95ReMM SLERP 13B$0.35$0.656,144Jul 22, 2023
liquidLiquidAI: LFM2.5-24B-A2B$0.05$0.70128,000Oct 13, 2026
nousresearchNous: Hermes 3 70B Instruct$0.70$0.70131,072Aug 18, 2024
inceptionInception: Mercury 2$0.25$0.75128,000Mar 4, 2026
mancerMancer: Weaver (alpha)$0.40$0.758,000Aug 2, 2023
sao10kSao10K: Llama 3.3 Euryale 70B$0.65$0.75131,072Dec 18, 2024
arcee-aiArcee AI: Trinity Large Thinking$0.25$0.80262,144Apr 1, 2026
thedrummerTheDrummer: Skyfall 36B V2$0.55$0.8032,768Mar 10, 2025
sao10kSao10K: Llama 3.1 Euryale 70B v2.2$0.85$0.85131,072Aug 28, 2024
cognitivecomputationsVenice: Uncensored$0.20$0.90128,000Jul 9, 2025
nousresearchNous: Hermes 3 405B Instruct$1.00$1.00131,072Aug 16, 2024
stepfunStepFun: Step 3.7 Flash$0.20$1.15262,144May 28, 2026
meituanMeituan: LongCat 2.0$0.30$1.201,048,756Jul 20, 2026
thinkingmachinesThinking Machines: Inkling Small$0.45$1.20524,288Jul 30, 2026
morphMorph: Morph V3 Fast$0.80$1.2081,920Jul 7, 2025
relaceRelace: Relace Apply 3$0.85$1.25256,000Sep 26, 2025
perceptronPerceptron: Perceptron Mk1$0.15$1.5032,768May 12, 2026
perceptronPerceptron: Perceptron Mk1.5$0.15$1.5036,864Sep 25, 2026
morphMorph: Morph V3 Large$0.90$1.90262,144Jul 7, 2025
moondreamMoondream: Moondream 3 Preview$0.30$2.5032,000Oct 15, 2026
moondreamMoondream: Moondream 3.1$0.30$2.5032,000Oct 15, 2026
stepfunStepFun: Step 5 Preview$1.00$2.701,048,576Oct 8, 2026
elorianElorian: Glimpse$0.60$3.00262,144Oct 17, 2026
nousresearchNous: Hermes 4 405B$1.00$3.00131,072Aug 26, 2025
reflectionReflection: Beam 1.0 501B A23B$1.00$3.001,048,576Oct 16, 2026
relaceRelace: Relace Search$1.00$3.00256,000Dec 8, 2025
unbiasedPareto 26.10$0.80$3.201,048,576Oct 15, 2026
sakanaSakana: Sakana Namazu$0.95$4.00262,144Aug 11, 2026
thinkingmachinesThinking Machines: Inkling$1.00$4.05524,288Jul 17, 2026
anthracite-orgMagnum v4 72B$2.50$5.0032,768Oct 22, 2024
writerWriter: Palmyra X5$0.60$6.001,040,000Jan 21, 2026
sakanaSakana: Fugu Max$2.00$6.001,000,000Sep 11, 2026
unbiasedPareto 26.9$2.50$7.50262,144Oct 13, 2026
fireworksFireworks: Ember-1$3.00$15.001,048,576Sep 24, 2026
sakanaSakana: Fugu Ultra$5.00$30.001,000,000Jun 24, 2026
sakanaSakana: Fugu Ultra v2$5.00$30.001,000,000Sep 11, 2026

Free variants (7)

$0 per token on the latest check. A free variant can be slower, capped or withdrawn, so free variants are listed apart from paid prices.

ProviderModelInput per 1M tokensOutput per 1M tokensContext (tokens)Listed since
dots-studioDots Studio: Dots3-Note Preview (free)$0.00$0.00512,000Aug 14, 2026
liquidLiquidAI: LFM2.5-2.6B (free)$0.00$0.0065,536Aug 11, 2026
poolsidePoolside: Laguna S 2.1 (free)$0.00$0.00262,144Jul 21, 2026
poolsidePoolside: Laguna XS 2.1 (free)$0.00$0.00262,144Jul 2, 2026
thinkingmachinesThinking Machines: Inkling (free)$0.00$0.001,048,576Jul 17, 2026
thinkingmachinesThinking Machines: Inkling Small (free)$0.00$0.001,048,576Jul 30, 2026
xingchen-agiXingChen: Xing4.0 29B A4B (free)$0.00$0.00262,144Oct 16, 2026

No longer listed: last listed prices (6)

These models are not on OpenRouter’s current list in our latest checks; each row shows the last price listed, with its date, not a current price.

ProviderModelInput per 1M tokensOutput per 1M tokensContext (tokens)Listed since
apodexApodex: Apodex 1.1 Mini (free) (last listed on OpenRouter Oct 13, 2026)$0.00$0.00262,144Oct 1, 2026
stealthSpace Bunny Alpha (last listed on OpenRouter Oct 5, 2026)$0.00$0.001,000,000Sep 23, 2026
baiduBaidu: ERNIE 4.5 VL 424B A47B (last listed on OpenRouter Oct 8, 2026)$0.42$1.25123,000Jun 30, 2025
kwaipilotKwaipilot: KAT-Coder-Pro V2.5 (last listed on OpenRouter Oct 7, 2026)$0.74$2.96262,144Jul 10, 2026
unbiasedPareto 26.10 Preview (last listed on OpenRouter Oct 15, 2026)$0.80$3.201,048,576Oct 1, 2026
unbiasedPareto (last listed on OpenRouter Oct 13, 2026)$2.50$7.50262,144Sep 17, 2026

Unknown means no value was recorded at our latest check; it is never shown as zero. Prices in USD as OpenRouter lists them; we do not add taxes.

How these numbers are produced

Each row is one model at one provider in our ai-models section. Prices are the input and output prices per one million tokens from the latest check of that model; the context length is the value read on the same pass. The table is sorted by our latest output price, lowest first; rows without an output price come last. The order is by price only.

A provider gets its own page when we track at least 5 of its models.

Each model links to its page with dates, the source link and the full history.

Methodology

What changed

Listed prices on OpenRouter move often. Confirmed changes in AI model prices, subscription limits and terms are listed day by day on What changed in AI models, prices and terms; every model page keeps its own history.

Caliper Press AI API Price Watch

$1.80 per 1M tokens

Median OpenRouter output price per 1M tokens across paid models listed on the date (free and batch variants excluded). based on 345 · . Methodology

345 records are in this figure; each has its own page with dates and sources.

Caliper Press AI API Price Watch — major-vendor flagships (direct vendor price)

$6.00 per 1M tokens

Median official direct output price per 1M tokens across flagship models of major vendors. based on 17 · Oct 18, 2026.

Caliper Press AI API Price Watch — major-vendor flagships (OpenRouter price)

$5.06 per 1M tokens

Median OpenRouter output price per 1M tokens across flagship models of major vendors. based on 22 · Oct 18, 2026.