API Cost Check
Cheapest LLM API by workload
Among 86 models released in the last year, the lowest input price is Claude Haiku 5.5 at $0.10 per 1M tokens and the lowest output price is MiMo V2.6 Flash at $0.28. The cheapest model for you depends on the mix of input, output and cached tokens, so each table below ranks models for one typical workload; put your own numbers into the calculator for an exact ranking.
Estimates from published per-token rates; check official pricing before budgeting. Data synced from the LiteLLM price list 2026-10-12 08:00 UTC.
Cheapest for: Customer support chatbot
5,000 req/day · 1,200 in · 400 out · cache 60% · Calculate with your numbers →
| # | Model | Provider | Monthly | Input /1M | Output /1M | Cached input /1M |
|---|---|---|---|---|---|---|
| 1 | MiMo V2.6 Flash | Xiaomi | $27.18/mo | $0.14 | $0.28 | $0.0028 |
| 2 | Claude Haiku 5.5 | Anthropic | $38.51/mo | $0.10 | $0.50 | $0.01 |
| 3 | GPT-6 Luna | OpenAI | $38.51/mo | $0.10 | $0.50 | $0.01 |
| 4 | Qwen3.8 Flash | Alibaba (Qwen) | $41.18/mo | $0.15 | $0.47 | $0.016 |
| 5 | GLM-5.3 Flash | Z.ai | $44.04/mo | $0.15 | $0.50 | $0.03 |
| 6 | Mistral Small (latest) | Mistral | $48.42/mo | $0.15 | $0.60 | $0.015 |
| 7 | Mercury 2.5 | Inception | $61.56/mo | $0.20 | $0.75 | $0.02 |
| 8 | Qwen3 VL 32b Instruct | Alibaba (Qwen) | $67.20/mo | $0.16 | $0.64 | — |
| 9 | Qwen3 VL 32b Thinking | Alibaba (Qwen) | $67.20/mo | $0.16 | $0.64 | — |
| 10 | MiMo V2.6 Pro | Xiaomi | $83.91/mo | $0.435 | $0.87 | $0.0036 |
Cheapest for: RAG over documents
2,000 req/day · 8,000 in · 600 out · cache 30% · Calculate with your numbers →
| # | Model | Provider | Monthly | Input /1M | Output /1M | Cached input /1M |
|---|---|---|---|---|---|---|
| 1 | Claude Haiku 5.5 | Anthropic | $53.64/mo | $0.10 | $0.50 | $0.01 |
| 2 | GPT-6 Luna | OpenAI | $53.64/mo | $0.10 | $0.50 | $0.01 |
| 3 | MiMo V2.6 Flash | Xiaomi | $57.52/mo | $0.14 | $0.28 | $0.0028 |
| 4 | Qwen3.8 Flash | Alibaba (Qwen) | $70.82/mo | $0.15 | $0.47 | $0.016 |
| 5 | GLM-5.3 Flash | Z.ai | $72.72/mo | $0.15 | $0.50 | $0.03 |
| 6 | Mistral Small (latest) | Mistral | $74.16/mo | $0.15 | $0.60 | $0.015 |
| 7 | Mercury 2.5 | Inception | $97.08/mo | $0.20 | $0.75 | $0.02 |
| 8 | Qwen3 VL 32b Instruct | Alibaba (Qwen) | $99.84/mo | $0.16 | $0.64 | — |
| 9 | Qwen3 VL 32b Thinking | Alibaba (Qwen) | $99.84/mo | $0.16 | $0.64 | — |
| 10 | GPT-5.6 Luna | OpenAI | $114/mo | $0.20 | $1.20 | $0.02 |
Cheapest for: Coding agent
1,500 req/day · 25,000 in · 2,000 out · cache 80% · Calculate with your numbers →
| # | Model | Provider | Monthly | Input /1M | Output /1M | Cached input /1M |
|---|---|---|---|---|---|---|
| 1 | MiMo V2.6 Flash | Xiaomi | $59.22/mo | $0.14 | $0.28 | $0.0028 |
| 2 | Claude Haiku 5.5 | Anthropic | $79.31/mo | $0.10 | $0.50 | $0.01 |
| 3 | GPT-6 Luna | OpenAI | $79.31/mo | $0.10 | $0.50 | $0.01 |
| 4 | Qwen3.8 Flash | Alibaba (Qwen) | $96.08/mo | $0.15 | $0.47 | $0.016 |
| 5 | Mistral Small (latest) | Mistral | $101/mo | $0.15 | $0.60 | $0.015 |
| 6 | GLM-5.3 Flash | Z.ai | $106/mo | $0.15 | $0.50 | $0.03 |
| 7 | Mercury 2.5 | Inception | $130/mo | $0.20 | $0.75 | $0.02 |
| 8 | GPT-5.4 nano | OpenAI | $176/mo | $0.20 | $1.25 | $0.02 |
| 9 | GPT-5.6 Luna | OpenAI | $177/mo | $0.20 | $1.20 | $0.02 |
| 10 | MiMo V2.6 Pro | Xiaomi | $179/mo | $0.435 | $0.87 | $0.0036 |
Cheapest for: Deep reasoning
800 req/day · 3,000 in · 6,000 out · cache 20% · Calculate with your numbers →
| # | Model | Provider | Monthly | Input /1M | Output /1M | Cached input /1M |
|---|---|---|---|---|---|---|
| 1 | MiMo V2.6 Flash | Xiaomi | $48.42/mo | $0.14 | $0.28 | $0.0028 |
| 2 | Qwen3.8 Flash | Alibaba (Qwen) | $76.55/mo | $0.15 | $0.47 | $0.016 |
| 3 | Claude Haiku 5.5 | Anthropic | $77.90/mo | $0.10 | $0.50 | $0.01 |
| 4 | GPT-6 Luna | OpenAI | $77.90/mo | $0.10 | $0.50 | $0.01 |
| 5 | GLM-5.3 Flash | Z.ai | $81.07/mo | $0.15 | $0.50 | $0.03 |
| 6 | Mistral Small (latest) | Mistral | $95.26/mo | $0.15 | $0.60 | $0.015 |
| 7 | Qwen3 VL 32b Instruct | Alibaba (Qwen) | $104/mo | $0.16 | $0.64 | — |
| 8 | Qwen3 VL 32b Thinking | Alibaba (Qwen) | $104/mo | $0.16 | $0.64 | — |
| 9 | Mercury 2.5 | Inception | $120/mo | $0.20 | $0.75 | $0.02 |
| 10 | MiMo V2.6 Pro | Xiaomi | $150/mo | $0.435 | $0.87 | $0.0036 |
Cheapest for: Batch processing
50,000 req/day · 1,500 in · 150 out · batch · Calculate with your numbers →
| # | Model | Provider | Monthly | Input /1M | Output /1M | Cached input /1M |
|---|---|---|---|---|---|---|
| 1 | Claude Haiku 5.5 | Anthropic | $169/mo | $0.10 | $0.50 | $0.01 |
| 2 | GPT-6 Luna | OpenAI | $169/mo | $0.10 | $0.50 | $0.01 |
| 3 | GPT-5.6 Luna | OpenAI | $360/mo | $0.20 | $1.20 | $0.02 |
| 4 | GPT-5.4 nano | OpenAI | $366/mo | $0.20 | $1.25 | $0.02 |
| 5 | MiMo V2.6 Flash | Xiaomi | $378/mo | $0.14 | $0.28 | $0.0028 |
| 6 | Qwen3.8 Flash | Alibaba (Qwen) | $443/mo | $0.15 | $0.47 | $0.016 |
| 7 | GLM-5.3 Flash | Z.ai | $450/mo | $0.15 | $0.50 | $0.03 |
| 8 | Gemini 3.1 Flash Lite | $450/mo | $0.25 | $1.50 | $0.025 | |
| 9 | Mistral Small (latest) | Mistral | $472/mo | $0.15 | $0.60 | $0.015 |
| 10 | Qwen3 VL 32b Instruct | Alibaba (Qwen) | $504/mo | $0.16 | $0.64 | — |
Before you pick the cheapest
A cheaper model that needs retries, longer prompts or a bigger model as fallback can cost more per completed task, so test two candidates on your own prompts and compare quality per dollar. Discounts change the order: see the chatbot and RAG guides for caching, and Batch API pricing for offline work.
FAQ
Which LLM API is the cheapest overall?
There is no single answer: input, output and cached tokens are priced separately, so a model that wins on input can lose on output. The tables above rank models for each typical workload; the first paragraph names the lowest input and output prices in our data.
Are free tiers included?
No. Rankings use paid, pay-as-you-go prices only. Free tiers, credits, committed-use discounts and regional prices are not included.
How often is this list updated?
Every 4 hours, from the same price data as the rest of the site; the sync time is shown above the tables. Price changes are logged on the price changes page, and you can get alerts per model.