Cheapest LLM APIs (Live Cost Ranking)
For high-volume classification, routing, RAG extraction, and agent loops, token price dominates unit economics. Below are the lowest-cost paid API models currently tracked, sorted live by blended 3:1 cost.
| # | Model | Provider | Input / 1M | Output / 1M | Blended | Context |
|---|---|---|---|---|---|---|
| 1 |
inclusionAI: Ling 3.0 Tiny
ultra-budget
|
Inclusionai | $0.015 | $0.045 | $0.022 | 262K |
| 2 |
Meta: Llama 3.1 8B Instruct
Open Weight
|
Meta Llama | $0.02 | $0.04 | $0.025 | 131K |
| 3 |
Mistral: Mistral Nemo
Open Weight
|
Mistral AI | $0.029 | $0.03 | $0.029 | 131K |
| 4 |
ScaleDown: Extract
Vision
|
Scaledown | $0.04 | Free | $0.03 | 1M |
| 5 |
ScaleDown: Summarize
Vision
|
Scaledown | $0.04 | Free | $0.03 | 1M |
| 6 |
inclusionAI: Ling 3.0 Flash
ultra-budget
|
Inclusionai | $0.021 | $0.063 | $0.032 | 262K |
| 7 |
OpenAI: gpt-oss-20b
ultra-budget
|
OpenAI | $0.017 | $0.085 | $0.034 | 131K |
| 8 |
Sao10K: Llama 3 8B Lunaris
ultra-budget
|
Sao10k | $0.04 | $0.05 | $0.043 | 8K |
| 9 |
Nex AGI: Nex-N2.5-Mini
Vision
|
Nex Agi | $0.025 | $0.10 | $0.044 | 262K |
| 10 |
Z.ai: GLM 5.3 Flash
Vision
|
Z Ai | $0.04 | $0.078 | $0.05 | 1.05M |
| 11 |
Z.ai: GLM Flash Latest
Vision
|
Z Ai | $0.04 | $0.078 | $0.05 | 1.05M |
| 12 |
Qwen: Qwen3.7 Flash
ReasoningVisionOpen Weight
|
Qwen | $0.03 | $0.13 | $0.055 | 1M |
| 13 |
Mistral: Mistral Small 3
ultra-budget
|
Mistral AI | $0.05 | $0.08 | $0.058 | 33K |
| 14 |
Inference.net: Schematron V2 Turbo
ultra-budget
|
Inference Net | $0.03 | $0.15 | $0.06 | 128K |
| 15 |
Amazon: Nova Micro 1.0
ultra-budget
|
Amazon | $0.035 | $0.14 | $0.061 | 128K |
| 16 |
Tencent: Hy4 flash
Vision
|
Tencent | $0.035 | $0.141 | $0.061 | 1.05M |
| 17 |
Google: Gemma 3 4B
VisionOpen Weight
|
$0.05 | $0.10 | $0.063 | 131K | |
| 18 |
Cohere: Command R7B (12-2024)
ultra-budget
|
Cohere | $0.037 | $0.15 | $0.066 | 128K |
| 19 |
DeepSeek: DeepSeek V4 Flash Latest
Open Weight
|
DeepSeek | $0.044 | $0.132 | $0.066 | 1.05M |
| 20 |
Google: Gemma 3 12B
VisionOpen Weight
|
$0.05 | $0.15 | $0.075 | 131K | |
| 21 |
Tencent: Hy-MT2-1.8B
ultra-budget
|
Tencent | $0.044 | $0.177 | $0.077 | 8K |
| 22 |
IBM: Granite 4.2 8B
Reasoning
|
Ibm Granite | $0.06 | $0.15 | $0.083 | 131K |
| 23 |
DeepSeek: DeepSeek V4.1 Flash
VisionOpen Weight
|
DeepSeek | $0.05 | $0.199 | $0.087 | 1.05M |
| 24 |
Upstage: Solar Mini 4
ultra-budget
|
Upstage | $0.05 | $0.20 | $0.087 | 524K |
| 25 |
Microsoft: Phi 4
Open Weight
|
Microsoft | $0.07 | $0.14 | $0.087 | 16K |
Try These Models Without Multiple Billing Accounts
OpenRouter
Unified API gateway for 300+ LLMs with automatic fallbacks and unified billing.
Together AI
High-throughput serverless inference & fine-tuning for Llama, DeepSeek, and Qwen.
GroqCloud
LPU-accelerated inference delivering hundreds of tokens per second for real-time voice & agents.
Some outbound infrastructure and API gateway links on KostoZero are partner referrals that help keep this observatory free and independent at no extra cost to you.
Get Instant Price-Drop (>10%), New AI Bot & SaaS Cliff Alerts
Every 6 hours our autonomous sync checks 40+ LLMs, AI crawlers, and SaaS pricing cliffs—alerting you when prices drop >10% or new scrapers appear.