Models
Cheapest LLM APIs for Developers
Ranked leaderboard of LLM APIs by estimated list price for a fixed sample workload — find the lowest-cost models before you commit. For OpenAI-only cuts, see Reduce OpenAI API Cost; for leaving OpenAI, see Alternatives.
By AI Credits Editorial · Content updated 2026-08-28
What this ranking measures
Models are sorted by estimated cost for 1M input + 500K output tokens at public list rates. The ranking answers “which model is cheapest on paper for this mix?” — not which is best for coding, vision, or long-context agents.
How to use the list
Shortlist the top few models that support your required features (tools, JSON mode, context window), then run quality and latency evals. Re-check prices after vendor announcements — list rates move, and relays can undercut official prices with different reliability trade-offs.
Related pages with different jobs
Token Costs explains billing units. LLM Cost Optimization covers routing and caching tactics. Cheaper OpenAI Alternatives compares non-OpenAI options against GPT-4o mini specifically. Calculators estimate a custom traffic mix.
Ranked for 1M input + 500K output tokens at public list price.
- 1
OpenAI · GPT-4o mini
$0.150/1M in · $0.600/1M out
$0.450
Free credits → - 2
Google · Gemini 2.5 Flash
$0.150/1M in · $0.600/1M out
Free tier available via AI Studio
$0.450
Free credits → - 3
Anthropic · Claude Haiku
$0.250/1M in · $1.25/1M out
$0.875
Free credits → - 4
Google · Gemini 2.5 Pro
$1.25/1M in · $10.00/1M out
$6.25
Free credits → - 5
OpenAI · GPT-4o
$2.50/1M in · $10.00/1M out
$7.50
Free credits → - 6
Anthropic · Claude Sonnet
$3.00/1M in · $15.00/1M out
$10.50
Free credits →
FAQ
Frequently asked questions
Quick answers about credits, eligibility, and how we verify offers.