Guides · Updated 2026-10-21
LLM APIs Under $1 per Million Tokens
Under $1/1M input you get serious capability: modern open-weight models that handle chat, extraction, classification, and even light coding. This is the sweet spot for high-volume production workloads.
💡 Current leader: IBM: Granite 4.0 Micro at $0.017 input / $0.112 output per 1M tokens.
Ranking (40 models)
| # | Model | Input /1M | Output /1M | Context |
|---|---|---|---|---|
| 1 | IBM: Granite 4.0 Micro ibm-granite | $0.017 | $0.112 | 131K |
| 2 | OpenAI: gpt-oss-20b openai | $0.018 | $0.090 | 131K |
| 3 | inclusionAI: Ling 3.0 Flash inclusionai | $0.021 | $0.063 | 262K |
| 4 | OpenAI: gpt-oss-20b (batch) openai | $0.024 | $0.112 | 131K |
| 5 | Nex AGI: Nex-N2.5-Mini nex-agi | $0.025 | $0.100 | 262K |
| 6 | OpenAI: GPT-5 Nano (batch) openai | $0.025 | $0.200 | 400K |
| 7 | Meta: Llama 3.2 1B Instruct meta-llama | $0.027 | $0.201 | 60K |
| 8 | Mistral: Mistral Nemo mistralai | $0.029 | $0.030 | 131K |
| 9 | OpenAI: gpt-oss-120b (batch) openai | $0.030 | $0.136 | 131K |
| 10 | Inference.net: Schematron V2 Turbo inference-net | $0.030 | $0.150 | 128K |
| 11 | Qwen: Qwen3.7 Flash qwen | $0.030 | $0.130 | 1M |
| 12 | Z.ai: GLM 5.2 z-ai | $0.034 | $7.00 | 1.0M |
| 13 | Amazon: Nova Micro 1.0 amazon | $0.035 | $0.140 | 128K |
| 14 | Cohere: Command R7B (12-2024) cohere | $0.037 | $0.150 | 128K |
| 15 | ScaleDown: Summarize scaledown | $0.040 | Free | 1M |
| 16 | ScaleDown: Extract scaledown | $0.040 | Free | 1M |
| 17 | Inception: Mercury 2.5 inception | $0.040 | $0.150 | 260K |
| 18 | Qwen: Qwen3.8 27B qwen | $0.040 | $1.35 | 1M |
| 19 | Google: Gemma 3 4B google | $0.040 | $0.050 | 131K |
| 20 | Sao10K: Llama 3 8B Lunaris sao10k | $0.040 | $0.050 | 8K |
| 21 | Tencent: Hy-MT2-1.8B tencent | $0.044 | $0.177 | 8K |
| 22 | DeepSeek: DeepSeek V4 Flash Latest ~deepseek | $0.048 | $0.157 | 1.0M |
| 23 | DeepSeek: DeepSeek Flash Latest ~deepseek | $0.049 | $0.375 | 1.0M |
| 24 | DeepSeek: DeepSeek V4.1 Flash deepseek | $0.049 | $0.375 | 1.0M |
| 25 | OpenAI: GPT-5 Nano openai | $0.050 | $0.400 | 400K |
| 26 | LiquidAI: LFM2.5-24B-A2B liquid | $0.050 | $0.700 | 128K |
| 27 | Anthropic: Claude Haiku 5.5 (batch) anthropic | $0.050 | $0.250 | 1M |
| 28 | Upstage: Solar Mini 4 upstage | $0.050 | $0.200 | 524K |
| 29 | OpenAI: GPT-6 Luna Pro (batch) openai | $0.050 | $0.250 | 1.1M |
| 30 | OpenAI: GPT-6 Luna (batch) openai | $0.050 | $0.250 | 1.1M |
| 31 | Inference.net: Schematron V2 Small inference-net | $0.050 | $0.230 | 128K |
| 32 | DeepSeek: DeepSeek V4.1 Flash (batch) deepseek | $0.050 | $0.150 | 1.0M |
| 33 | Z.ai: GLM 5.3 Flash (batch) z-ai | $0.050 | $0.180 | 1.0M |
| 34 | Google: Gemini 2.5 Flash Lite (batch) google | $0.050 | $0.200 | 1.0M |
| 35 | OpenAI: GPT-4.1 Nano (batch) openai | $0.050 | $0.200 | 1.0M |
| 36 | Google: Gemma 3 12B google | $0.050 | $0.150 | 131K |
| 37 | Mistral: Mistral Small 3 mistralai | $0.050 | $0.080 | 33K |
| 38 | Meta: Llama 3.1 8B Instruct meta-llama | $0.050 | $0.080 | 131K |
| 39 | Meta: Llama 3.2 3B Instruct meta-llama | $0.051 | $0.335 | 131K |
| 40 | Z.ai: GLM Flash Latest ~z-ai | $0.052 | $0.120 | 1.0M |
Prices are list prices via OpenRouter, verified 2026-10-21. Provider first-party pricing may differ (volume discounts, batch tiers).
FAQ
- Are sub-$1 LLM APIs good enough for production?
- For many workloads, yes. Modern open-weight models in this range handle RAG, classification, extraction, summarization, and chat competently. Reserve frontier models for tasks where quality directly drives revenue.
- What are the tradeoffs of cheap models?
- Typically smaller context windows, weaker reasoning on hard problems, and less reliable instruction-following. Check the context column below — some budget models offer surprisingly large windows.
- How much would 10M tokens per month cost?
- At $0.30/1M input, roughly $3 per 10M input tokens plus output costs. Most teams in this tier spend single-digit dollars monthly.
Estimate your own bill with the cost calculator · Browse all 458 models · More buying guides