Monthly API cost of an AI support bot handling 3,000 tickets, compared across every major model.
Bottom line: cheapest is Meta: Llama 3.2 1B Instruct at about $1.03/month; the median model costs $25.14/month; the most expensive, Anthropic: Claude Fable 5, costs $335.16/month. That is a 325× spread for the same workload.
Scenario: 3,000 support tickets/month × 4 turns, 1,500 words (help-docs context + thread) in / 120 words out per turn.
| Vendor | Model | In $/1M | Out $/1M | Est. monthly cost | Source / check |
|---|---|---|---|---|---|
| Meta (Llama) | Meta: Llama 3.2 1B Instruct | $0.027 | $0.201 | $1.03 | source 2026-10-21 ✔ matches |
| Meta (Llama) | Meta: Llama 3.1 8B Instruct | $0.05 | $0.08 | $1.35 | source 2026-10-21 ℹ vs market median ($0.05/$0.07) |
| DeepSeek | DeepSeek: DeepSeek V4.1 Flash | $0.0489 | $0.1956 | $1.55 | source 2026-10-21 ⚠ differs from lab price ($0.14/$0.28) |
| Meta (Llama) | Meta: Llama 3.2 3B Instruct | $0.0509 | $0.335 | $1.86 | source 2026-10-21 ✔ matches |
| DeepSeek | DeepSeek: DeepSeek V4 Flash 0423 | $0.091 | $0.182 | $2.53 | source 2026-10-21 ℹ vs market median ($0.14/$0.28) |
| Mistral | Mistral: Ministral 3 3B 2512 | $0.1 | $0.1 | $2.59 | source 2026-10-21 ✔ matches |
| Meta (Llama) | Meta: Llama 4 Scout | $0.1 | $0.3 | $2.97 | source 2026-10-21 ✔ matches |
| Meta (Llama) | Meta: Llama 3.3 70B Instruct | $0.1 | $0.32 | $3.01 | source 2026-10-21 ℹ vs market median ($0.5683/$0.72) |
| Anthropic | Anthropic: Claude Haiku 5.5 | $0.1 | $0.5 | $3.35 | source 2026-10-21 ✔ matches |
| OpenAI | OpenAI: GPT-6 Luna Pro | $0.1 | $0.5 | $3.35 | source 2026-10-21 ✔ matches |
| OpenAI | OpenAI: GPT-6 Luna | $0.1 | $0.5 | $3.35 | source 2026-10-21 ✔ matches |
| Mistral | Mistral: Ministral 3 8B 2512 | $0.15 | $0.15 | $3.88 | source 2026-10-21 ✔ matches |
| Meta (Llama) | Meta: Llama Guard 4 12B | $0.18 | $0.18 | $4.65 | source 2026-10-21 ✔ matches |
| DeepSeek | DeepSeek: DeepSeek V4.1 Flash 1017 | $0.15 | $0.6 | $4.74 | source 2026-10-21 no 2nd source |
| Mistral | Mistral: Mistral Small 4 | $0.15 | $0.6 | $4.74 | source 2026-10-21 ✔ matches |
| Mistral | Mistral: Ministral 3 14B 2512 | $0.2 | $0.2 | $5.17 | source 2026-10-21 ✔ matches |
| Meta (Llama) | Meta: Llama 4 Maverick | $0.1875 | $0.6525 | $5.74 | source 2026-10-21 ℹ vs market median ($0.22/$0.8381) |
| DeepSeek | DeepSeek: DeepSeek V4 Flash Vision Exp | $0.2156 | $0.6468 | $6.40 | source 2026-10-21 ℹ vs market median ($0.2246/$0.6739) |
| DeepSeek | DeepSeek: DeepSeek V3.2 | $0.28 | $0.42 | $7.51 | source 2026-10-21 ✔ matches |
| Google: Gemini 3.1 Flash Lite | $0.25 | $1.5 | $8.86 | source 2026-10-21 ✔ matches | |
| Meta (Llama) | Meta: Llama 3.1 70B Instruct | $0.4 | $0.4 | $10.34 | source 2026-10-21 ℹ vs market median ($0.4/$0.56) |
| Google: Gemini 3.5 Flash Lite | $0.3 | $2.5 | $11.97 | source 2026-10-21 ✔ matches | |
| DeepSeek | DeepSeek: DeepSeek V4 Pro 0423 | $0.4461 | $0.8923 | $12.39 | source 2026-10-21 ℹ vs market median ($1.5008/$3.2) |
| DeepSeek | DeepSeek: DeepSeek V4 Flash 0731 | $0.44 | $1.32 | $13.06 | source 2026-10-21 ℹ vs market median ($0.143/$0.4) |
| Mistral | Mistral: Devstral 2 2512 | $0.4 | $2.0 | $13.41 | source 2026-10-21 ✔ matches |
| Mistral | Mistral: Mistral Large 3 2512 | $0.5 | $1.5 | $14.84 | source 2026-10-21 ✔ matches |
| DeepSeek | DeepSeek: DeepSeek V4 Pro 0813 | $0.66 | $1.98 | $19.59 | source 2026-10-21 ✔ matches |
| Google: Gemini 3.7 Flash | $0.75 | $3.75 | $25.14 | source 2026-10-21 no 2nd source | |
| Google: Gemini 3.6 Flash | $0.75 | $3.75 | $25.14 | source 2026-10-21 ✔ matches | |
| Google: Gemini 3.8 Flash | $0.825 | $4.125 | $27.65 | source 2026-10-21 ⚠ differs from lab price ($0.75/$3.75) | |
| xAI | SpaceXAI: Grok Build 0.1 | $1.0 | $2.0 | $27.77 | source 2026-10-21 ✔ matches |
| xAI | SpaceXAI: Grok 4.3 | $1.25 | $2.5 | $34.71 | source 2026-10-21 ✔ matches |
| xAI | SpaceXAI: Grok 4.20 Multi-Agent | $1.25 | $2.5 | $34.71 | source 2026-10-21 ✔ matches |
| xAI | SpaceXAI: Grok 4.20 | $1.25 | $2.5 | $34.71 | source 2026-10-21 ✔ matches |
| Mistral | Mistral: Mistral Large 4 | $1.36 | $4.18 | $40.56 | source 2026-10-21 ⚠ differs from lab price ($0.68/$2.09) |
| Google: Nano Banana 2.1 | $1.5 | $7.5 | $50.27 | source 2026-10-21 no 2nd source | |
| Mistral | Mistral: Mistral Medium 3.5 | $1.5 | $7.5 | $50.27 | source 2026-10-21 ✔ matches |
| Google: Gemini 3.5 Flash | $1.5 | $9.0 | $53.15 | source 2026-10-21 no 2nd source | |
| xAI | SpaceXAI: Grok 4.7 | $2.0 | $6.0 | $59.37 | source 2026-10-21 ✔ matches |
| xAI | SpaceXAI: Grok 4.6 | $2.0 | $6.0 | $59.37 | source 2026-10-21 ✔ matches |
| xAI | SpaceXAI: Grok 4.5 | $2.0 | $6.0 | $59.37 | source 2026-10-21 ✔ matches |
| Google: Gemini 4 Argon | $2.0 | $10.0 | $67.03 | source 2026-10-21 no 2nd source | |
| OpenAI | OpenAI: GPT-6.1 Sol Pro | $2.0 | $10.0 | $67.03 | source 2026-10-21 ✔ matches |
| OpenAI | OpenAI: GPT-6.1 Sol | $2.0 | $10.0 | $67.03 | source 2026-10-21 ✔ matches |
| Anthropic | Anthropic: Claude Sonnet 5.5 | $2.0 | $10.0 | $67.03 | source 2026-10-21 ✔ matches |
| OpenAI | OpenAI: GPT-6 Sol Pro | $2.0 | $10.0 | $67.03 | source 2026-10-21 ✔ matches |
| OpenAI | OpenAI: GPT-6 Sol | $2.0 | $10.0 | $67.03 | source 2026-10-21 ✔ matches |
| Anthropic | Anthropic: Claude Sonnet 5 | $2.0 | $10.0 | $67.03 | source 2026-10-21 ✔ matches |
| Anthropic | Anthropic: Claude Opus 5.5 | $4.0 | $20.0 | $134.06 | source 2026-10-21 ✔ matches |
| Anthropic | Anthropic: Claude Opus 5 | $5.0 | $25.0 | $167.58 | source 2026-10-21 ✔ matches |
| Anthropic | Anthropic: Claude Opus 4.8 | $5.0 | $25.0 | $167.58 | source 2026-10-21 ✔ matches |
| OpenAI | OpenAI: GPT-6 Astra | $10.0 | $50.0 | $335.16 | source 2026-10-21 ✔ matches |
| OpenAI | OpenAI: GPT-6 Astra Pro | $10.0 | $50.0 | $335.16 | source 2026-10-21 ✔ matches |
| Anthropic | Anthropic: Claude Fable 5.1 | $10.0 | $50.0 | $335.16 | source 2026-10-21 ✔ matches |
| Anthropic | Anthropic: Claude Fable 5 | $10.0 | $50.0 | $335.16 | source 2026-10-21 ✔ matches |
Try the interactive calculators with your own numbers →
Every week: which AI API prices changed, and what it means for your app's bill. Free.
Prices from OpenRouter's public API, cross-checked with InferIndex. No spam.