LLM pricing

SAKANA AI · PERFORMANCE-MULTI-AGENT

Fugu Ultra by Sakana AI

$5 input / $30 output per million tokens. 272K context window. Full pricing hub live.

$5/M in$30/M out272K

OPENAI · FRONTIER

GPT-5.6 Sol by OpenAI

$5 input / $30 output per million tokens. 1.05M context window. Full pricing hub live.

$5/M in$30/M out1.05M

GOOGLE · FRONTIER-FLASH

Gemini 3.5 Flash by Google

$1.50 input / $9 output per million tokens. 1M context window. Full pricing hub live.

$1.50/M in$9/M out1M

AI21 LABS · FRONTIER

Jamba Large 1.7 by AI21 Labs

$2 input / $8 output per million tokens. 256K context window. Full pricing hub live.

$2/M in$8/M out256K

§ 02 / PROVIDERS

Browse by provider.

All models →

OPEN-WEIGHT GIANT

Alibaba (Qwen)

The Qwen (Tongyi Qianwen) model family from Alibaba Cloud's Tongyi Lab. First released in 2023, Qwen is the most-downloaded open-weight model family in the world — most tiers ship under Apache 2.0 on Hugging Face and ModelScope, while the proprietary Max tier is served through Alibaba Cloud's Model Studio.

28 models · from $0.05/M →

CHINA CONSUMER LEADER

ByteDance (Doubao)

The Doubao model family from ByteDance — parent of TikTok and Douyin. Served through the Volcano Ark (火山方舟) platform on ByteDance's Volcano Engine cloud, Doubao powers China's most-used consumer AI app and undercuts most frontier labs on price.

14 models · from $0.02/M →

WECHAT-SCALE LAB

Tencent (Hunyuan)

The Hunyuan (混元) model family from Tencent — parent of WeChat, QQ, and the world's largest games business. Served through Tencent Cloud and the Yuanbao assistant inside WeChat, Hunyuan reaches over a billion users through China's dominant super-app.

14 models · from $0.07/M →

GLM ARCHITECT

Zhipu (Z.ai / GLM)

The Beijing lab behind the GLM models, also operating internationally as Z.ai. Spun out of Tsinghua University in 2019, Zhipu ships strong coding/agentic models, multiple free tiers, and open weights — and was the first of China's "Six Tigers" to pursue an IPO. It also sits on the US Entity List.

14 models · from $0.07/M →

OpenAI

The lab behind ChatGPT and the GPT family. Founded as a nonprofit in 2015 by Sam Altman, Elon Musk, Ilya Sutskever, Greg Brockman and others; restructured into a capped-profit LLC in 2019 and finally into a public-benefit corporation on October 28, 2025. Ships the largest commercial model lineup, the largest consumer AI product, and — after the April 2026 funding round — the highest-valued private AI company in history at ~$852B.

13 models · from $0.02/M →

EU FRONTIER LAB

Mistral AI

The Paris lab behind Le Chat and the open-weight Mistral family. Founded in 2023 by Arthur Mensch with ex-DeepMind and Meta researchers, it ships frontier-grade models with Apache-2.0 open weights for most tiers and EU-native data residency — the leading European alternative to US labs.

12 models · from $0.1/M →

MEDICAL-AI SPECIALIST

Baichuan

The Beijing lab behind the Baichuan general models and the medical M-series. Founded in 2023 by Wang Xiaochuan (former CEO of search engine Sogou), Baichuan has pivoted hard toward healthcare AI — its open-source Baichuan-M3 tops several medical benchmarks — while keeping a general-purpose 4-series lineup.

10 models · from $0.07/M →

Anthropic

The AI safety-focused lab behind Claude. Founded by ex-OpenAI researchers in 2021, it ships frontier-grade reasoning models with zero-retention defaults and the most rigorous published RSP (Responsible Scaling Policy) in the field.

9 models · from $1/M →

AGENTIC + MEDIA LAB

MiniMax

The Shanghai lab behind the agentic MiniMax M2 models and the Hailuo video/audio generators. Founded in 2021 by Yan Junjie (a former SenseTime research lead), MiniMax pairs cheap long-context text models with consumer media apps — and listed in Hong Kong in January 2026.

9 models · from $0.3/M →

xAI

Elon Musk's AI lab, founded March 2023 as X.AI Corp. Ships the Grok family — frontier reasoning and coding models trained on the Memphis Colossus supercluster (150 MW peak, plans for 1M+ GPUs). Acquired by SpaceX on Feb 2, 2026 at a $250B xAI valuation ($1.25T combined entity). Brand still operates as xAI; API still at docs.x.ai; Grok still integrated natively into X (Twitter).

6 models · from $1/M →

Google

Google DeepMind is the AI division of Alphabet, formed in April 2023 by merging the original DeepMind (founded London, 2010, acquired by Google in 2014) with Google Brain. CEO Demis Hassabis. Ships the Gemini family — the most multimodal frontier lineup on the market (text + image + audio + video native) — through both the Gemini API and the cloud platform now rebranded as Gemini Enterprise Agent Platform (formerly Vertex AI, renamed April 2026).

5 models · from $0.1/M →

ENTERPRISE LLM

Cohere

Cohere is an enterprise AI company behind Command, Aya, Embed, Rerank, and North. Public token pricing is currently explicit for Aya Expanse; newer Command models are listed as live but use custom or private-deployment pricing.

4 models · from $0.04/M →

WEB-GROUNDED

Perplexity

Perplexity's API centers on web-grounded answers. Sonar models combine token billing with search-context request fees, so the full cost depends on both generated tokens and retrieval depth.

4 models · from $1/M →

OPEN-WEIGHT FRONTIER

Moonshot (Kimi)

The lab behind Kimi — the long-context assistant — and the open-weight Kimi K2 models. Founded in 2023 by Yang Zhilin with Tsinghua schoolmates, Moonshot ships a 1-trillion-parameter open-weight MoE that rivals frontier US labs while staying downloadable on Hugging Face.

3 models · from $0.6/M →

MULTIMODAL CHAT

Reka

Reka is an AI lab offering Chat, Research, and Vision APIs. For AI//COST we track the public text-token rows for Reka Edge, Reka Flash, and Reka Core.

3 models · from $0.1/M →

EU / ISRAEL LAB

AI21 Labs

Jamba is the model family from AI21 Labs, an Israeli AI lab founded in 2017. Jamba pairs a Mamba state-space architecture with Transformer layers for a 256K-token context at a lower cost than comparable long-context models, and ships as Jamba Large and Jamba Mini.

2 models · from $0.2/M →

CHINA FRONTIER

Baidu

ERNIE (Wenxin) is the large-language-model family built by Baidu, China's largest search company, and served through its Qianfan MaaS platform. The lineup runs from the ERNIE 5.x frontier tier down to the cost-efficient ERNIE 4.5 Turbo, all priced per token in Chinese yuan.

2 models · from $0.12/M →

COST DISRUPTOR

DeepSeek

The Chinese lab behind the DeepSeek open-weight models. Spun out of quant hedge fund High-Flyer in 2023 by Liang Wenfeng, it shocked the industry by matching frontier reasoning quality at a fraction of the cost — and shipping the weights under an MIT license.

2 models · from $0.14/M →

CLOUD GIANT

Amazon

Amazon Nova is the family of first-party foundation models built by Amazon and served on Amazon Bedrock. The current Nova 2 generation replaced the original Nova Micro/Lite/Pro/Premier lineup, led by Nova 2 Lite at a 1M-token context window with on-demand per-token pricing.

1 model · from $0.3/M →

MULTI-AGENT MODEL

Sakana AI

Sakana AI describes Sakana Fugu as a multi-agent system delivered as one model API. Fugu dynamically orchestrates a pool of expert models for coding, reasoning, research, security analysis, and other complex multi-step work.

1 model · from $5/M →

§ 02 / HEAD-TO-HEADS

Most-read comparisons.

All comparisons →

ChatGPT Free vs ChatGPT Paid

ChatGPT Free runs GPT-5.5 Instant with capped messages, may show ads in the US, and marks most features 'limited' — no GPT-5.6 reasoning mod…

OpenAI vs OpenAI↗

ChatGPT Go vs ChatGPT Plus

ChatGPT Go ($8/mo) is Free with roughly 10× the headroom — more messages, uploads, and image generation — but it keeps GPT-5.5 Instant only,…

OpenAI vs OpenAI↗

ChatGPT Plus vs ChatGPT Pro

ChatGPT Plus ($20/mo) and Pro ($200/mo) run the same models, but Pro unlocks the deepest reasoning — GPT-5.6 Sol Pro plus Extra High — 20× t…

OpenAI vs OpenAI↗

Claude Code vs Cursor

Two different shapes of tool. Claude Code is a terminal-native agent on Anthropic's models; Cursor is an AI-native IDE that runs Claude, GPT…

Anthropic vs Anysphere↗

Claude Free vs Claude Pro

Since June 30, 2026 Free and Pro run the same default model, Sonnet 5 — so raw answer quality is identical. Pro's $20/mo buys roughly 5× the…

Anthropic vs Anthropic↗

Claude Sonnet 5 · Scheduled move to standard Sonnet pricing: $3/M input and $15/M output

Claude Pro vs Claude Max

Same Claude, same Opus 4.8 — the only difference is how much you can run. Pro is $20/mo and enough for most; Max is $100 (5×) or $200 (20×) …

Anthropic vs Anthropic↗

§ 03 / SCOREBOARD

Live pricing leaders.

Methodology →

Metric	What it measures	Current leader	Top value	Models tracked	Last updated
Cheapest input	Lowest $/M input tokens — for high-volume RAG and bulk classification.	Doubao Seed 1.6 Flash	$0.02/M	256	today
Cheapest output	Lowest $/M output tokens — matters most for long-form generation.	Ministral 3B	$0.04/M	255	today
Biggest cache discount	Largest gap between fresh-input and cached-input pricing — best for agentic workloads.	MiMo-V2.5-Pro	−99%	142	today
Longest context	Largest context window — best for big-document analysis at high input cost.	Gemini 2.5 Pro	2M	260	today
Most expensive output	Highest $/M output — usually the frontier tier, used selectively.	GPT-5.4 Pro	$180/M	260	today

The tape doesn't lie.

Every price change, cache rollout, and model release in the LLM cluster — logged within hours, not weeks.

SEP 01

Jamba Large 1.7 · Verified Jamba Large 1.7 unchanged at $2/M input and $8/M output on AI21 Studio

Jamba Mini 1.7 · Verified Jamba Mini 1.7 unchanged at $0.20/M input and $0.40/M output on AI21 Studio

Amazon Nova 2 Lite · Verified global cross-region pricing at $0.30/M input and $2.50/M output

ERNIE 4.5 Turbo · Verified on Qianfan billing at CNY 0.8/M input, CNY 3.2/M output, CNY 0.2/M cache-hit

ERNIE 5.1 · Verified on Qianfan billing at CNY 4/M input and CNY 18/M output (0-32K tier)

Doubao Seed 1.6 Vision · Base <=32K input and output doubled to Y0.80/Y8.00 (cache-hit input Y0.16 unchanged)

Command R+ · Verified public Cohere token pricing unchanged

§ 04 / BEST FOR

Pick by your workload.

If you need frontier reasoning + tool use with cache enabled…

Claude Fable 5

If you're optimising for cost per million tokens in bulk RAG…

Doubao Seed 1.6 Flash

If long context matters more than price…

Gemini 2.5 Pro

If you want balance of cost and capability for a chatbot…

Claude Sonnet 4.6

If the workload is simple classification at scale…

Claude Haiku 4.5