Models / Cheapest

The cheapest LLM APIs

Every generally-available model, ranked by blended price per 1M tokens. Sortable and filterable — find the cheapest option that clears your quality bar.

The short answer

The cheapest generally-available LLM API we track is Llama 3.1 8B Instruct (Meta) at $0.022 per 1M tokens blended — $0.02 input / $0.03 output. Across the 62 priced GA chat models we track, blended cost ranges from $0.022 to $67.50 per 1M tokens.

Last verified: 9 Jul 2026 · sourced from official provider documentation

Cheapest overallLlama 3.1 8B Instruct
$0.022/1M blended · $0.02 in / $0.03 out · 128K context · Meta
The ranking

62 models, cheapest first

Ranked by blended $/1M tokens. Click a header to re-sort, or filter by provider.

Model Status Context Input $/M Output $/M Blended $/M Cutoff
Llama 3.1 8B Instruct
Meta · text
GA128K$0.02$0.03$0.0222023-12
Amazon Nova Micro
Amazon · text
GA128K$0.035$0.14$0.0612024-10
Command R7B
Cohere · text
GA128K$0.037$0.15$0.066
Ministral 3 3B
Mistral · text, vision
GA$0.1$0.1$0.1
Amazon Nova Lite
Amazon · text, image, video
GA300K$0.06$0.24$0.1052024-10
Qwen-Flash
Alibaba · text
GA1M$0.05$0.4$0.138
Llama 4 Scout (17B-16E Instruct)
Meta · text, image
GA10M$0.1$0.3$0.152024-08
Ministral 3 8B
Mistral · text, vision
GA256K$0.15$0.15$0.15
Llama 3.3 70B Instruct
Meta · text
GA128K$0.1$0.32$0.1552023-12
Qwen3.5-Flash
Alibaba · text, image
GA1M$0.1$0.4$0.175
Ministral 3 14B
Mistral · text, vision
GA$0.2$0.2$0.2
Llama 4 Maverick (17B-128E Instruct)
Meta · text, image
GA1M$0.15$0.6$0.2622024-08
Mistral Small 4
Mistral · text, vision
GA256K$0.15$0.6$0.262
Command R (08-2024)
Cohere · text
GA128K$0.15$0.6$0.262
Codestral (v25.08)
Mistral · text, code
GA128K$0.3$0.9$0.45
GPT-5.4 nano
OpenAI · text, image
GA400K$0.2$1.25$0.4632025-08-31
Gemini 3.1 Flash-Lite
Google · text, image, video, audio, pdf
GA1.0M$0.25$1.50$0.5632025-01
Qwen3.6-Flash
Alibaba · text, image, video
GA1M$0.25$1.50$0.563
Qwen-Plus (Qwen3-series)
Alibaba · text
GA1M$0.4$1.20$0.6
Moonshot v1 8K
Moonshot · text
GA8K$0.2$2$0.65
Qwen3.7-Plus
Alibaba · text, image, video
GA1M$0.4$1.60$0.7
Qwen3.7-Plus (snapshot 2026-05-26)
Alibaba · text, image, video
GA1M$0.4$1.60$0.7
Mistral Large 3
Mistral · text, vision
GA256K$0.5$1.50$0.75
Amazon Nova 2 Lite
Amazon · text, image, video
GA1M$0.3$2.50$0.852025-10
Qwen3.5-Plus
Alibaba · text, image
GA262K$0.4$2.40$0.9
Sonar
Perplexity · text, image
GA128K$1$1$1
Qwen3.6-Plus
Alibaba · text, image, video
GA1M$0.5$3$1.13
Kimi K2.5
Moonshot · text, image, video
GA262K$0.6$3$1.20
Grok Build 0.1
xAI · text-in, image-in, text-out
GA256K$1$2$1.25
Amazon Nova Pro
Amazon · text, image, video
GA300K$0.8$3.20$1.402024-10
Moonshot v1 32K
Moonshot · text
GA33K$1$3$1.50
Grok 4.3
xAI · text-in, image-in, text-out
GA1M$1.25$2.50$1.56
Grok 4.20 (0309) Reasoning
xAI · text-in, image-in, text-out
GA1M$1.25$2.50$1.56
Grok 4.20 (0309) Non-Reasoning
xAI · text-in, image-in, text-out
GA1M$1.25$2.50$1.56
GPT-5.4 mini
OpenAI · text, image
GA400K$0.75$4.50$1.692025-08-31
Kimi K2.7 Code
Moonshot · text
GA262K$0.95$4$1.71
Kimi K2.6
Moonshot · text, image, video
GA262K$0.95$4$1.71
Claude Haiku 4.5
Anthropic · text, image, vision
GA200K$1$5$22025-02
Qwen3-Max
Alibaba · text
GA262K$1.20$6$2.402025-06
Moonshot v1 128K
Moonshot · text
GA131K$2$5$2.75
Qwen-Max (Qwen2.5-Max)
Alibaba · text
GA33K$1.60$6.40$2.80
Mistral Medium 3.5
Mistral · text, vision
GA$1.50$7.50$3
Gemini 3.5 Flash
Google · text, image, video, audio, pdf
GA1.0M$1.50$9$3.382025-01
Kimi K2.7 Code HighSpeed
Moonshot · text
GA262K$1.90$8$3.42
Sonar Reasoning Pro
Perplexity · text
GA128K$2$8$3.50
Qwen3.7-Max
Alibaba · text, image, video
GA1M$2.50$7.50$3.75
Command A
Cohere · text
GA256K$2.50$10$4.38
Command R+ (08-2024)
Cohere · text
GA128K$2.50$10$4.38
GPT-5.3-Codex
OpenAI · text, image
GA400K$1.75$14$4.812025-08-31
GPT-5.4
OpenAI · text, image
GA1.1M$2.50$15$5.632025-08-31
Claude Sonnet 5
Anthropic · text, image, vision
GA1M$3$15$62026-01
Claude Sonnet 4.6
Anthropic · text, image, vision
GA1M$3$15$62025-08
Claude Sonnet 4.5
Anthropic · text, image, vision
GA200K$3$15$62025-01
Sonar Pro
Perplexity · text, image
GA200K$3$15$6
Claude Opus 4.8
Anthropic · text, image, vision
GA1M$5$25$102026-01
Claude Opus 4.7
Anthropic · text, image, vision
GA1M$5$25$102026-01
Claude Opus 4.6
Anthropic · text, image, vision
GA1M$5$25$102025-05
Claude Opus 4.5
Anthropic · text, image, vision
GA200K$5$25$102025-05
GPT-5.5
OpenAI · text, image
GA1.1M$5$30$11.252025-12-01
Claude Fable 5
Anthropic · text, image, vision
GA1M$10$50$20
GPT-5.5 Pro
OpenAI · text, image
GA1.1M$30$180$67.502025-12-01
GPT-5.4 Pro
OpenAI · text, image
GA1.1M$30$180$67.502025-08-31

Blended = 0.75 × input + 0.25 × output $/M tokens (a fair single-number cost proxy). Click any header to sort.

FAQ

Cheapest LLM APIs

What is the cheapest LLM API right now?

Llama 3.1 8B Instruct (Meta) is the cheapest generally-available LLM we track, at $0.02 per 1M input tokens and $0.03 per 1M output tokens — a blended $0.022 per 1M. The next cheapest is Amazon Nova Micro (Amazon) at a blended $0.061 per 1M. Prices are sourced from official provider documentation and refreshed daily.

How is the cheapest model ranked?

We rank every generally-available chat LLM by a blended price of 0.75 × input + 0.25 × output (USD per 1M tokens) — a single-number cost proxy that weights the input-heavy nature of most real workloads. Embeddings, rerank and media-generation models are excluded so the leaderboard is a like-for-like comparison. Click any column header to re-sort by input, output or context.

Is the cheapest LLM good enough for production?

Not always. Price is only one axis — quality, latency, rate limits and reliability matter just as much, and the cheapest model on paper can cost more once you account for retries or a larger prompt. Use this ranking to build a shortlist, then test the top few on your own workload before committing.

Do these prices include free tiers or discounts?

No. The blended figure is standard on-demand API pricing. Prompt caching, batch processing and provider free tiers can cut the real bill substantially — use the cost calculator to model your own token volumes and see how caching changes the ranking.