Independent and dated
What the AI APIs actually charge, read at the source
Side-by-side comparisons and cost calculators. Every figure comes from the provider's own pricing page and carries the day it was read, including the cache lines and context thresholds most comparisons skip.
- Pricing pages published
- 70
- Ship a calculator you can run
- 9
- Last price check
- 11 Nov 2026
Calculators
Run the numbers on your own volume
Calculator
LLM cost calculator
Describe a workload, read the invoice each model would send for it.
Open it
Calculator
OpenAI cost calculator
Fresh input, cached input and output, counted the way OpenAI counts them.
Open it
Calculator
Azure against the direct API
What the detour through Azure adds, or does not add, to the same deployment.
Open it
Calculator
Cheapest API, ranked
Metered APIs sorted by what a million tokens really costs today.
Open it
Calculator
Transcription per hour
Whisper, Deepgram, Groq and AssemblyAI, brought back to an hour of audio.
Open it
Start here
The reference table
Cost comparison
LLM costs per month
Output tokens, not input, drive almost every API bill you will ever receive. The spread between the cheapest and the most expensive frontier model is over sixtyfold: on 12 September 2026, DeepSeek V4.1-Flash charged…
Read the comparison
All comparisons
The full index, newest first
- Claude Mythos 5.2 pricing: $8 in, $40 out per million tokens Model pricing
- Gemini 4 Carbon: what it does and what it costs, $2 in and $10 out Model pricing
- Chutes vs OpenRouter: what the cheaper router actually costs API routers
- Claude Fable 5.2: pricing, benchmarks and what people build with it Model pricing
- Azure gpt-4o retires 14 January 2027: what gpt-5.1 costs instead Model pricing
- GPT Image 2.6 pricing: $8 in, $30 out per million tokens Model pricing
- GPT-6 Cyber pricing: $20 in, $100 out per million tokens Model pricing
- Gemini Omni 1.2 Flash pricing: $0.07 per second of 720p video Model pricing
- Groq free tier limits: what you get before you pay Model pricing
- Claude Ultra’s $200 API credits: what they buy in tokens Model pricing
- GPT-6.1 Terra pricing: $1 in, $5 out per million tokens Model pricing
- Groq’s Llama models went Contact Sales: what replaces the $0.05 tier Model pricing
- Kimi K3.1 Pro pricing: $2 in, $10 out per million tokens Model pricing
- Inception Mercury 3 pricing: $0.20 in, $0.70 out per million Model pricing
- Eleven v4 Turbo pricing: half the v4 rate for real-time speech Audio APIs
- Voxtral 3 transcription pricing, per hour of audio Audio APIs
- Eleven Realtime API pricing: per minute plus tokens Audio APIs
- Voxtral 3 TTS pricing: $13 per million characters Audio APIs
- DeepSeek Search and Fetch API pricing: $0.75 per 1,000 requests Model pricing
- Eleven v4 pricing: what an hour of narration costs per character Audio APIs
- Claude Sonnet 5.5 pricing: the halved cache read and what it saves Model pricing
- GPT-6.1 Sol pricing: what the halved cache rate changes Model pricing
- OpenAI Decisions API pricing: $0.10 per million input tokens and what is not billed Model pricing
- When gpt-5.6-sol stops being cheap: budgeting past 30 November 2026 Model pricing
- GPT-6.1 Sol Ultrafast pricing: what 8x speed costs per million tokens Model pricing
- Mistral Large 4 pricing: $0.70 in, $2.10 out after the open-weight price cut Model pricing
- Gemini 4 Argon: what it does and what it costs, $2 in and $10 out until 30 November Model pricing
- Claude Haiku 5.5 pricing: the 100K-token threshold that decides your bill Model pricing
- Grok Voice Transcribe 2.0 pricing, per hour of audio Audio APIs
- Grok 4.7 pricing: what xAI charges per million tokens Model pricing
Explore
Browse by category
Model pricing
What each AI model actually charges, provider by provider, with the date of every price check.
47 comparisons
Audio APIs
Speech, transcription and voice APIs, priced per hour of real audio.
9 comparisons
API routers
Routers and gateways that sit between you and the model providers, and what they add to the bill.
6 comparisons
Cost comparison
Side-by-side cost of running the same workload across providers.
5 comparisons
AI girlfriend apps
Companion and chat apps, with what the subscription really costs once the add-ons are counted.
3 comparisons