The cheapest LLM API per million tokens, ranked

The cheapest LLM API per million tokens is not a single model — it depends on how your traffic splits between input and output — but the budget field clusters tightly at the bottom, and one legacy model still undercuts the whole current generation. This is that tier, ranked, with the fine print that decides the real winner.

To rank models that price input and output differently on one axis, we use a blended price: (3 × input + 1 × output) ÷ 4, a rough 3-to-1 input-to-output mix typical of RAG and chat. Your real ratio changes the order — an output-heavy workload punishes high output rates far more — so treat the ranking as a starting point and price your own mix.

The budget tier, ranked by blended price

Blended prices are rounded to the cent. The three-way tie at $0.18 — Gemini 2.5 Flash-Lite, DeepSeek V4 Flash off-peak, and Qwen-Flash — splits differently depending on your ratio: DeepSeek has the lowest output rate ($0.28), so it pulls ahead on output-heavy work, while Gemini and Qwen have the lowest input rate ($0.10) and win on input-heavy work.

The cheapest overall: GPT-5 nano, with an asterisk

On raw price, legacy GPT-5 nano wins outright at $0.05 / $0.40 — a blended $0.14, cheaper than anything in the current generation. The asterisk is "legacy": it is a prior-generation model that a provider can deprecate with notice, so building a long-lived product on it carries migration risk. Among current-generation, generally-available budget models, the real floor is the $0.18 cluster: Gemini 2.5 Flash-Lite, DeepSeek V4 Flash (off-peak), and Qwen-Flash.

What that means in dollars

Take a high-volume workload of 10M input and 2M output tokens per month. GPT-5 nano runs about $1.30; Gemini 2.5 Flash-Lite and Qwen-Flash about $1.80; DeepSeek V4 Flash off-peak about $1.96. The same workload on GPT-5.4 mini is about $16.50, and on Claude Haiku 4.5 about $20.00 — a 10-to-15× spread inside what everyone calls the "budget" tier. Picking the wrong budget model costs more than picking a flagship for a task that needed one.

DeepSeek has the cheapest cached input anywhere

If your workload reuses a large, stable prompt prefix, DeepSeek V4 Flash is in a class of its own: cached input lists at about $0.0028 per 1M tokens — a ~98% discount off its $0.14 input rate, and the cheapest cached read in the dataset. The catch is the same off-peak caveat: during peak hours the rate doubles. For a cache-heavy pipeline scheduled off-peak, nothing else comes close.

The caveats that decide the real winner

• ◦Legacy status: GPT-5 nano is a prior-generation model — cheapest on sticker, but subject to deprecation. Weigh migration risk for anything long-lived.
• ◦Off-peak vs peak: DeepSeek's $0.14 / $0.28 (and its ~$0.0028 cached rate) are off-peak; the peak rate (09:00–12:00 and 14:00–18:00 Beijing time, UTC+8) is 2×.
• ◦Third-party hosting: Meta does not sell Llama tokens directly — Llama 4 Scout's price is Together AI's serverless rate, and other hosts price it differently.
• ◦Promotions: Qwen's budget rates can include limited-time promos; verify the current price before you rely on it.
• ◦Audio surcharge: Gemini Flash-Lite models price audio input higher than text — Gemini 2.5 Flash-Lite text input is $0.10, but audio runs higher. Check the modality you actually send.
• ◦Cost per task, not per token: the cheapest per-token model is not automatically the cheapest per task — output length and tokenizer differences can flip the ranking. Model your real prompts.

Which should you pick

For most teams the honest answer is one of three:

• ◦Absolute lowest cost, short-lived or internal: legacy GPT-5 nano ($0.05 / $0.40), accepting deprecation risk.
• ◦Lowest cost on a current, GA model: Gemini 2.5 Flash-Lite or Qwen-Flash for input-heavy work; DeepSeek V4 Flash off-peak for output-heavy or cache-heavy work.
• ◦Lowest cost with no off-peak clock and a US-hosted API: GPT-5.6 Luna ($0.20 / $1.20) or Mistral Small 4 ($0.15 / $0.60).

What is the cheapest LLM API right now?

By blended price, legacy GPT-5 nano ($0.14) is cheapest overall; among current-generation GA models, Gemini 2.5 Flash-Lite, DeepSeek V4 Flash off-peak, and Qwen-Flash tie at about $0.18. Which of the three is cheapest for you depends on whether your workload is input- or output-heavy.

Is a cheaper model always the right choice?

No. A budget model that needs three attempts or a longer prompt to match a workhorse's single-shot answer can cost more in total, and it may miss on quality, latency, or tool use. Cheapest per token is a starting filter, not the decision.

How do I compute the cost for my own workload?

Take your monthly input and output token volumes, multiply each by the model's per-1M rate, and add them — then layer in cache-hit rate and any off-peak or long-context surcharge. The pricing calculator does this for any of the 40 models and lets you compare two side by side.

Do these prices change?

Frequently. These are list prices as of 2026-08-03; budget pricing is the most volatile tier, and promos and off-peak windows come and go. Re-verify with each provider before committing.